From 544160b6ea18670196d1173c099f2cced5075132 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 24 Jan 2017 15:18:24 -0800 Subject: [PATCH 001/119] fbdev: color map copying bounds checking commit 2dc705a9930b4806250fbf5a76e55266e59389f2 upstream. Copying color maps to userspace doesn't check the value of to->start, which will cause kernel heap buffer OOB read due to signedness wraps. CVE-2016-8405 Link: http://lkml.kernel.org/r/20170105224249.GA50925@beast Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Kees Cook Reported-by: Peter Pi (@heisecode) of Trend Micro Cc: Min Chong Cc: Dan Carpenter Cc: Tomi Valkeinen Cc: Bartlomiej Zolnierkiewicz Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- drivers/video/fbdev/core/fbcmap.c | 26 ++++++++++++++------------ 1 file changed, 14 insertions(+), 12 deletions(-) diff --git a/drivers/video/fbdev/core/fbcmap.c b/drivers/video/fbdev/core/fbcmap.c index f89245b8ba8e..68a113594808 100644 --- a/drivers/video/fbdev/core/fbcmap.c +++ b/drivers/video/fbdev/core/fbcmap.c @@ -163,17 +163,18 @@ void fb_dealloc_cmap(struct fb_cmap *cmap) int fb_copy_cmap(const struct fb_cmap *from, struct fb_cmap *to) { - int tooff = 0, fromoff = 0; - int size; + unsigned int tooff = 0, fromoff = 0; + size_t size; if (to->start > from->start) fromoff = to->start - from->start; else tooff = from->start - to->start; - size = to->len - tooff; - if (size > (int) (from->len - fromoff)) - size = from->len - fromoff; - if (size <= 0) + if (fromoff >= from->len || tooff >= to->len) + return -EINVAL; + + size = min_t(size_t, to->len - tooff, from->len - fromoff); + if (size == 0) return -EINVAL; size *= sizeof(u16); @@ -187,17 +188,18 @@ int fb_copy_cmap(const struct fb_cmap *from, struct fb_cmap *to) int fb_cmap_to_user(const struct fb_cmap *from, struct fb_cmap_user *to) { - int tooff = 0, fromoff = 0; - int size; + unsigned int tooff = 0, fromoff = 0; + size_t size; if (to->start > from->start) fromoff = to->start - from->start; else tooff = from->start - to->start; - size = to->len - tooff; - if (size > (int) (from->len - fromoff)) - size = from->len - fromoff; - if (size <= 0) + if (fromoff >= from->len || tooff >= to->len) + return -EINVAL; + + size = min_t(size_t, to->len - tooff, from->len - fromoff); + if (size == 0) return -EINVAL; size *= sizeof(u16); From e4be4d4942b92cdd2d3e87ed178269bcd5844526 Mon Sep 17 00:00:00 2001 From: Dave Martin Date: Fri, 6 Jan 2017 17:54:51 +0000 Subject: [PATCH 002/119] tile/ptrace: Preserve previous registers for short regset write commit fd7c99142d77dc4a851879a66715abf12a3193fb upstream. Ensure that if userspace supplies insufficient data to PTRACE_SETREGSET to fill all the registers, the thread's old registers are preserved. Signed-off-by: Dave Martin Signed-off-by: Chris Metcalf Signed-off-by: Greg Kroah-Hartman --- arch/tile/kernel/ptrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/tile/kernel/ptrace.c b/arch/tile/kernel/ptrace.c index d89b7011667c..e279572824b1 100644 --- a/arch/tile/kernel/ptrace.c +++ b/arch/tile/kernel/ptrace.c @@ -111,7 +111,7 @@ static int tile_gpr_set(struct task_struct *target, const void *kbuf, const void __user *ubuf) { int ret; - struct pt_regs regs; + struct pt_regs regs = *task_pt_regs(target); ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, ®s, 0, sizeof(regs)); From 2abb7f408f7cfb9af9218e74507f5f44af154302 Mon Sep 17 00:00:00 2001 From: Peter Ujfalusi Date: Mon, 9 Jan 2017 16:31:58 +0200 Subject: [PATCH 003/119] drm: Schedule the output_poll_work with 1s delay if we have delayed event commit 68f458eec7069d618a6c884ca007426e0cea411b upstream. Instead of scheduling the work to handle the initial delayed event, use 1s delay. This delay should not be needed, but Optimus/nouveau will fail in a mysterious way if the delayed event is handled as soon as possible like it is done in drm_helper_probe_single_connector_modes() in case the poll was enabled before. Reverting 339fd36238dd would give back the 10 sec (!) delay to handle the delayed event. Adding 1sec delay to the poll_work is enough to work around the issue in Optimus setups and gives shorter response on handling the initial delayed event. Fixes: 339fd36238dd ("drm: drm_probe_helper: Fix output_poll_work scheduling") Signed-off-by: Peter Ujfalusi [danvet: Add FIXME to the comment to make it stick out more.] Signed-off-by: Daniel Vetter Link: http://patchwork.freedesktop.org/patch/msgid/20170109143158.21917-1-peter.ujfalusi@ti.com Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/drm_probe_helper.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c index f6b64d7d3528..276474d13763 100644 --- a/drivers/gpu/drm/drm_probe_helper.c +++ b/drivers/gpu/drm/drm_probe_helper.c @@ -143,8 +143,18 @@ void drm_kms_helper_poll_enable_locked(struct drm_device *dev) } if (dev->mode_config.delayed_event) { + /* + * FIXME: + * + * Use short (1s) delay to handle the initial delayed event. + * This delay should not be needed, but Optimus/nouveau will + * fail in a mysterious way if the delayed event is handled as + * soon as possible like it is done in + * drm_helper_probe_single_connector_modes() in case the poll + * was enabled before. + */ poll = true; - delay = 0; + delay = HZ; } if (poll) From f1dc9aaee02950e95cc045209d9bf07955bc16d8 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 9 Jan 2017 15:56:14 +0100 Subject: [PATCH 004/119] drm: Fix broken VT switch with video=1366x768 option MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit fdf35a6b22247746a7053fc764d04218a9306f82 upstream. I noticed that the VT switch doesn't work any longer with a Dell laptop with 1366x768 eDP when the machine is connected with a DP monitor. It behaves as if VT were switched, but the graphics remain frozen. Actually the keyboard works, so I could switch back to VT7 again. I tried to track down the problem, and encountered a long story until we reach to this error: - The machine is booted with video=1366x768 option (the distro installer seems to add it as default). - Recently, drm_helper_probe_single_connector_modes() deals with cmdline modes, and it tries to create a new mode when no matching mode is found. - The drm_mode_create_from_cmdline_mode() creates a mode based on either CVT of GFT according to the given cmdline mode; in our case, it's 1366x768. - Since both CVT and GFT can't express the width 1366 due to alignment, the resultant mode becomes 1368x768, slightly larger than the given size. - Later on, the atomic commit is performed, and in drm_atomic_check_only(), the size of each plane is checked. - The size check of 1366x768 fails due to the above, and eventually the whole VT switch fails. Back in the history, we've had a manual fix-up of 1368x768 in various places via c09dedb7a50e ("drm/edid: Add a workaround for 1366x768 HD panel"), but they have been all in drm_edid.c at probing the modes from EDID. For addressing the problem above, we need a similar hack to the mode newly created from cmdline, manually adjusting the width when the expected size is 1366 while we get 1368 instead. Fixes: eaf99c749d43 ("drm: Perform cmdline mode parsing during...") Signed-off-by: Takashi Iwai Link: http://patchwork.freedesktop.org/patch/msgid/20170109145614.29454-1-tiwai@suse.de Reviewed-by: Ville Syrjälä Signed-off-by: Ville Syrjälä Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/drm_modes.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c index 53f07ac7c174..e14366de0e6e 100644 --- a/drivers/gpu/drm/drm_modes.c +++ b/drivers/gpu/drm/drm_modes.c @@ -1462,6 +1462,13 @@ drm_mode_create_from_cmdline_mode(struct drm_device *dev, return NULL; mode->type |= DRM_MODE_TYPE_USERDEF; + /* fix up 1368x768: GFT/CVT can't express 1366 width due to alignment */ + if (cmd->xres == 1366 && mode->hdisplay == 1368) { + mode->hdisplay = 1366; + mode->hsync_start--; + mode->hsync_end--; + drm_mode_set_name(mode); + } drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V); return mode; } From 4c741e2adb351a44d1f4bc210aafd382939ac7ae Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Mon, 7 Nov 2016 22:20:54 +0200 Subject: [PATCH 005/119] drm/i915: Ignore bogus plane coordinates on SKL when the plane is not visible MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit 3bfdfdcbce2796ce75bf2d85fd8471858d702e5d upstream. When the plane is invisible we may have all sorts of bogus stuff in the coordinates, which we must ignore or else we might fail the plane update. This started to happen on SKL when I moved the plane offset computation to happen in the check phase. Previously we happily ignored it all since we never called the update_plane hook with an invisible plane. Cc: Sivakumar Thulasimani Cc: drm-intel-fixes@lists.freedesktop.org Fixes: b63a16f6cd89 ("drm/i915: Compute display surface offset in the plane check hook for SKL+") Signed-off-by: Ville Syrjälä Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98258 Testcase: igt/pm_rpm/legacy-planes Testcase: igt/pm_rpm/universal-planes Reviewed-by: Matt Roper Signed-off-by: Matt Roper Link: http://patchwork.freedesktop.org/patch/msgid/1478550057-24864-3-git-send-email-ville.syrjala@linux.intel.com (cherry picked from commit a5e4c7d0aa6784d8abe95c3ceef0da9656d17468) Signed-off-by: Jani Nikula Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/intel_display.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 869b29fe9ec4..3c9cb35e46ab 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2975,6 +2975,9 @@ int skl_check_plane_surface(struct intel_plane_state *plane_state) unsigned int rotation = plane_state->base.rotation; int ret; + if (!plane_state->base.visible) + return 0; + /* Rotate src coordinates to match rotated GTT view */ if (intel_rotation_90_or_270(rotation)) drm_rect_rotate(&plane_state->base.src, From 32600835ebe1bd972d01ce7f331b87bd80705d30 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Mon, 10 Oct 2016 09:44:06 -0700 Subject: [PATCH 006/119] drm/vc4: Fix memory leak of the CRTC state. commit 7622b25543665567d8830a63210385b7d705924b upstream. The underscores variant frees the pointers inside, while the no-underscores variant calls underscores and then frees the struct. Signed-off-by: Eric Anholt Fixes: d8dbf44f13b9 ("drm/vc4: Make the CRTCs cooperate on allocating display lists.") Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/vc4/vc4_crtc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index 7f08d681a74b..d544ff9b0d46 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -832,7 +832,7 @@ static void vc4_crtc_destroy_state(struct drm_crtc *crtc, } - __drm_atomic_helper_crtc_destroy_state(state); + drm_atomic_helper_crtc_destroy_state(crtc, state); } static const struct drm_crtc_funcs vc4_crtc_funcs = { From b9edac54cb85da589ca809bf8dcf86e5cd3f41c0 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Tue, 17 Jan 2017 21:42:53 +1100 Subject: [PATCH 007/119] drm/vc4: Fix an integer overflow in temporary allocation layout. commit 0f2ff82e11c86c05d051cae32b58226392d33bbf upstream. We copy the unvalidated ioctl arguments from the user into kernel temporary memory to run the validation from, to avoid a race where the user updates the unvalidate contents in between validating them and copying them into the validated BO. However, in setting up the layout of the kernel side, we failed to check one of the additions (the roundup() for shader_rec_offset) against integer overflow, allowing a nearly MAX_UINT value of bin_cl_size to cause us to under-allocate the temporary space that we then copy_from_user into. Reported-by: Murray McAllister Signed-off-by: Eric Anholt Fixes: d5b1a78a772f ("drm/vc4: Add support for drawing 3D frames.") Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/vc4/vc4_gem.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c index 303f23c96220..39ef6743d3b8 100644 --- a/drivers/gpu/drm/vc4/vc4_gem.c +++ b/drivers/gpu/drm/vc4/vc4_gem.c @@ -594,7 +594,8 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec) args->shader_rec_count); struct vc4_bo *bo; - if (uniforms_offset < shader_rec_offset || + if (shader_rec_offset < args->bin_cl_size || + uniforms_offset < shader_rec_offset || exec_size < uniforms_offset || args->shader_rec_count >= (UINT_MAX / sizeof(struct vc4_shader_state)) || From cfba2a001d0e36905016bb4f87fc47245c944c36 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Tue, 17 Jan 2017 21:58:06 +1100 Subject: [PATCH 008/119] drm/vc4: Return -EINVAL on the overflow checks failing. commit 6b8ac63847bc2f958dd93c09edc941a0118992d9 upstream. By failing to set the errno, we'd continue on to trying to set up the RCL, and then oops on trying to dereference the tile_bo that binning validation should have set up. Reported-by: Ingo Molnar Signed-off-by: Eric Anholt Fixes: d5b1a78a772f ("drm/vc4: Add support for drawing 3D frames.") Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/vc4/vc4_gem.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c index 39ef6743d3b8..18e37171e9c8 100644 --- a/drivers/gpu/drm/vc4/vc4_gem.c +++ b/drivers/gpu/drm/vc4/vc4_gem.c @@ -601,6 +601,7 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec) sizeof(struct vc4_shader_state)) || temp_size < exec_size) { DRM_ERROR("overflow in exec arguments\n"); + ret = -EINVAL; goto fail; } From 5270c017f19f32812d7d83ba6e45873b8db503d0 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 13 Jan 2017 10:49:00 +0300 Subject: [PATCH 009/119] drm/vc4: fix a bounds check commit 21ccc32496b2f63228f5232b3ac0e426e8fb3c31 upstream. We accidentally return success even if vc4_full_res_bounds_check() fails. Fixes: d5b1a78a772f ("drm/vc4: Add support for drawing 3D frames.") Signed-off-by: Dan Carpenter Reviewed-by: Eric Engestrom Reviewed-by: Eric Anholt Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/vc4/vc4_render_cl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/vc4/vc4_render_cl.c b/drivers/gpu/drm/vc4/vc4_render_cl.c index 08886a309757..5cdd003605f5 100644 --- a/drivers/gpu/drm/vc4/vc4_render_cl.c +++ b/drivers/gpu/drm/vc4/vc4_render_cl.c @@ -461,7 +461,7 @@ static int vc4_rcl_surface_setup(struct vc4_exec_info *exec, } ret = vc4_full_res_bounds_check(exec, *obj, surf); - if (!ret) + if (ret) return ret; return 0; From bbae3c4525966606db790ac7b7910b8639b5da42 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 25 Jan 2017 12:00:29 -0500 Subject: [PATCH 010/119] Revert "drm/radeon: always apply pci shutdown callbacks" commit b9b487e494712c8e5905b724e12f5ef17e9ae6f9 upstream. This seems to break reboot on some evergreen systems. bugs: https://bugs.freedesktop.org/show_bug.cgi?id=99524 https://bugzilla.kernel.org/show_bug.cgi?id=192271 This reverts commit a481daa88fd4d6b54f25348972bba10b5f6a84d0. Signed-off-by: Alex Deucher Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/radeon/radeon_drv.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index 00ea0002b539..e0c143b865f3 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -366,11 +366,10 @@ static void radeon_pci_shutdown(struct pci_dev *pdev) { /* if we are running in a VM, make sure the device - * torn down properly on reboot/shutdown. - * unfortunately we can't detect certain - * hypervisors so just do this all the time. + * torn down properly on reboot/shutdown */ - radeon_pci_remove(pdev); + if (radeon_device_is_virtual()) + radeon_pci_remove(pdev); } static int radeon_pmops_suspend(struct device *dev) From a2104c7cd3b24c4329b6193f7ec0882ce612f110 Mon Sep 17 00:00:00 2001 From: Lucas Stach Date: Fri, 27 Jan 2017 11:33:04 +0100 Subject: [PATCH 011/119] drm/atomic: clear out fence when duplicating state [Fixed differently in 4.10] The fence needs to be cleared out, otherwise the following commit might wait on a stale fence from the previous commit. This was fixed as a side effect of 9626014258a5 (drm/fence: add in-fences support) in kernel 4.10. As this commit introduces new functionality and as such can not be applied to stable, this patch is the minimal fix for the kernel 4.9 stable series. Signed-off-by: Lucas Stach Reviewed-by: Daniel Vetter Tested-by: Fabio Estevam Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/drm_atomic_helper.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c index 338766c64c99..a05bb3891119 100644 --- a/drivers/gpu/drm/drm_atomic_helper.c +++ b/drivers/gpu/drm/drm_atomic_helper.c @@ -3115,6 +3115,8 @@ void __drm_atomic_helper_plane_duplicate_state(struct drm_plane *plane, if (state->fb) drm_framebuffer_reference(state->fb); + + state->fence = NULL; } EXPORT_SYMBOL(__drm_atomic_helper_plane_duplicate_state); From 6676aa65464f23e32efe1857a51fe341dfb10e03 Mon Sep 17 00:00:00 2001 From: Keno Fischer Date: Tue, 24 Jan 2017 15:17:48 -0800 Subject: [PATCH 012/119] mm/huge_memory.c: respect FOLL_FORCE/FOLL_COW for thp commit 8310d48b125d19fcd9521d83b8293e63eb1646aa upstream. In commit 19be0eaffa3a ("mm: remove gup_flags FOLL_WRITE games from __get_user_pages()"), the mm code was changed from unsetting FOLL_WRITE after a COW was resolved to setting the (newly introduced) FOLL_COW instead. Simultaneously, the check in gup.c was updated to still allow writes with FOLL_FORCE set if FOLL_COW had also been set. However, a similar check in huge_memory.c was forgotten. As a result, remote memory writes to ro regions of memory backed by transparent huge pages cause an infinite loop in the kernel (handle_mm_fault sets FOLL_COW and returns 0 causing a retry, but follow_trans_huge_pmd bails out immidiately because `(flags & FOLL_WRITE) && !pmd_write(*pmd)` is true. While in this state the process is stil SIGKILLable, but little else works (e.g. no ptrace attach, no other signals). This is easily reproduced with the following code (assuming thp are set to always): #include #include #include #include #include #include #include #include #include #include #define TEST_SIZE 5 * 1024 * 1024 int main(void) { int status; pid_t child; int fd = open("/proc/self/mem", O_RDWR); void *addr = mmap(NULL, TEST_SIZE, PROT_READ, MAP_ANONYMOUS | MAP_PRIVATE, 0, 0); assert(addr != MAP_FAILED); pid_t parent_pid = getpid(); if ((child = fork()) == 0) { void *addr2 = mmap(NULL, TEST_SIZE, PROT_READ | PROT_WRITE, MAP_ANONYMOUS | MAP_PRIVATE, 0, 0); assert(addr2 != MAP_FAILED); memset(addr2, 'a', TEST_SIZE); pwrite(fd, addr2, TEST_SIZE, (uintptr_t)addr); return 0; } assert(child == waitpid(child, &status, 0)); assert(WIFEXITED(status) && WEXITSTATUS(status) == 0); return 0; } Fix this by updating follow_trans_huge_pmd in huge_memory.c analogously to the update in gup.c in the original commit. The same pattern exists in follow_devmap_pmd. However, we should not be able to reach that check with FOLL_COW set, so add WARN_ONCE to make sure we notice if we ever do. [akpm@linux-foundation.org: coding-style fixes] Link: http://lkml.kernel.org/r/20170106015025.GA38411@juliacomputing.com Signed-off-by: Keno Fischer Acked-by: Kirill A. Shutemov Cc: Greg Thelen Cc: Nicholas Piggin Cc: Willy Tarreau Cc: Oleg Nesterov Cc: Kees Cook Cc: Andy Lutomirski Cc: Michal Hocko Cc: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- mm/huge_memory.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 8ca40b70beae..917555cf6be0 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -772,6 +772,12 @@ struct page *follow_devmap_pmd(struct vm_area_struct *vma, unsigned long addr, assert_spin_locked(pmd_lockptr(mm, pmd)); + /* + * When we COW a devmap PMD entry, we split it into PTEs, so we should + * not be in this function with `flags & FOLL_COW` set. + */ + WARN_ONCE(flags & FOLL_COW, "mm: In follow_devmap_pmd with FOLL_COW set"); + if (flags & FOLL_WRITE && !pmd_write(*pmd)) return NULL; @@ -1118,6 +1124,16 @@ out_unlock: return ret; } +/* + * FOLL_FORCE can write to even unwritable pmd's, but only + * after we've gone through a COW cycle and they are dirty. + */ +static inline bool can_follow_write_pmd(pmd_t pmd, unsigned int flags) +{ + return pmd_write(pmd) || + ((flags & FOLL_FORCE) && (flags & FOLL_COW) && pmd_dirty(pmd)); +} + struct page *follow_trans_huge_pmd(struct vm_area_struct *vma, unsigned long addr, pmd_t *pmd, @@ -1128,7 +1144,7 @@ struct page *follow_trans_huge_pmd(struct vm_area_struct *vma, assert_spin_locked(pmd_lockptr(mm, pmd)); - if (flags & FOLL_WRITE && !pmd_write(*pmd)) + if (flags & FOLL_WRITE && !can_follow_write_pmd(*pmd, flags)) goto out; /* Avoid dumping huge zero page */ From 9b1a1ae9b5281e12893a34bcbd1686b5bcd2cd82 Mon Sep 17 00:00:00 2001 From: Vlastimil Babka Date: Tue, 24 Jan 2017 15:18:18 -0800 Subject: [PATCH 013/119] mm/mempolicy.c: do not put mempolicy before using its nodemask commit d51e9894d27492783fc6d1b489070b4ba66ce969 upstream. Since commit be97a41b291e ("mm/mempolicy.c: merge alloc_hugepage_vma to alloc_pages_vma") alloc_pages_vma() can potentially free a mempolicy by mpol_cond_put() before accessing the embedded nodemask by __alloc_pages_nodemask(). The commit log says it's so "we can use a single exit path within the function" but that's clearly wrong. We can still do that when doing mpol_cond_put() after the allocation attempt. Make sure the mempolicy is not freed prematurely, otherwise __alloc_pages_nodemask() can end up using a bogus nodemask, which could lead e.g. to premature OOM. Fixes: be97a41b291e ("mm/mempolicy.c: merge alloc_hugepage_vma to alloc_pages_vma") Link: http://lkml.kernel.org/r/20170118141124.8345-1-vbabka@suse.cz Signed-off-by: Vlastimil Babka Acked-by: Kirill A. Shutemov Acked-by: Michal Hocko Acked-by: David Rientjes Cc: Aneesh Kumar K.V Cc: Andrea Arcangeli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- mm/mempolicy.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/mempolicy.c b/mm/mempolicy.c index 0b859af06b87..f75704717e47 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -2023,8 +2023,8 @@ retry_cpuset: nmask = policy_nodemask(gfp, pol); zl = policy_zonelist(gfp, pol, node); - mpol_cond_put(pol); page = __alloc_pages_nodemask(gfp, order, zl, nmask); + mpol_cond_put(pol); out: if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie))) goto retry_cpuset; From ade7afe9dca6b13919f88abd38eefe32f22eaeb3 Mon Sep 17 00:00:00 2001 From: Vlastimil Babka Date: Tue, 24 Jan 2017 15:18:32 -0800 Subject: [PATCH 014/119] mm, page_alloc: fix check for NULL preferred_zone commit ea57485af8f4221312a5a95d63c382b45e7840dc upstream. Patch series "fix premature OOM regression in 4.7+ due to cpuset races". This is v2 of my attempt to fix the recent report based on LTP cpuset stress test [1]. The intention is to go to stable 4.9 LTSS with this, as triggering repeated OOMs is not nice. That's why the patches try to be not too intrusive. Unfortunately why investigating I found that modifying the testcase to use per-VMA policies instead of per-task policies will bring the OOM's back, but that seems to be much older and harder to fix problem. I have posted a RFC [2] but I believe that fixing the recent regressions has a higher priority. Longer-term we might try to think how to fix the cpuset mess in a better and less error prone way. I was for example very surprised to learn, that cpuset updates change not only task->mems_allowed, but also nodemask of mempolicies. Until now I expected the parameter to alloc_pages_nodemask() to be stable. I wonder why do we then treat cpusets specially in get_page_from_freelist() and distinguish HARDWALL etc, when there's unconditional intersection between mempolicy and cpuset. I would expect the nodemask adjustment for saving overhead in g_p_f(), but that clearly doesn't happen in the current form. So we have both crazy complexity and overhead, AFAICS. [1] https://lkml.kernel.org/r/CAFpQJXUq-JuEP=QPidy4p_=FN0rkH5Z-kfB4qBvsf6jMS87Edg@mail.gmail.com [2] https://lkml.kernel.org/r/7c459f26-13a6-a817-e508-b65b903a8378@suse.cz This patch (of 4): Since commit c33d6c06f60f ("mm, page_alloc: avoid looking up the first zone in a zonelist twice") we have a wrong check for NULL preferred_zone, which can theoretically happen due to concurrent cpuset modification. We check the zoneref pointer which is never NULL and we should check the zone pointer. Also document this in first_zones_zonelist() comment per Michal Hocko. Fixes: c33d6c06f60f ("mm, page_alloc: avoid looking up the first zone in a zonelist twice") Link: http://lkml.kernel.org/r/20170120103843.24587-2-vbabka@suse.cz Signed-off-by: Vlastimil Babka Acked-by: Mel Gorman Acked-by: Hillf Danton Cc: Ganapatrao Kulkarni Cc: Michal Hocko Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- include/linux/mmzone.h | 6 +++++- mm/page_alloc.c | 2 +- 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 0f088f3a2fed..f99c993dd500 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -972,12 +972,16 @@ static __always_inline struct zoneref *next_zones_zonelist(struct zoneref *z, * @zonelist - The zonelist to search for a suitable zone * @highest_zoneidx - The zone index of the highest zone to return * @nodes - An optional nodemask to filter the zonelist with - * @zone - The first suitable zone found is returned via this parameter + * @return - Zoneref pointer for the first suitable zone found (see below) * * This function returns the first zone at or below a given zone index that is * within the allowed nodemask. The zoneref returned is a cursor that can be * used to iterate the zonelist with next_zones_zonelist by advancing it by * one before calling. + * + * When no eligible zone is found, zoneref->zone is NULL (zoneref itself is + * never NULL). This may happen either genuinely, or due to concurrent nodemask + * update due to cpuset modification. */ static inline struct zoneref *first_zones_zonelist(struct zonelist *zonelist, enum zone_type highest_zoneidx, diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 34ada718ef47..593a11d8bc6b 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3763,7 +3763,7 @@ retry_cpuset: */ ac.preferred_zoneref = first_zones_zonelist(ac.zonelist, ac.high_zoneidx, ac.nodemask); - if (!ac.preferred_zoneref) { + if (!ac.preferred_zoneref->zone) { page = NULL; goto no_zone; } From d1656c5aef4d72f03a7833d07a378c8f604b8307 Mon Sep 17 00:00:00 2001 From: Vlastimil Babka Date: Tue, 24 Jan 2017 15:18:35 -0800 Subject: [PATCH 015/119] mm, page_alloc: fix fast-path race with cpuset update or removal commit 16096c25bf0ca5d87e4fa6ec6108ba53feead212 upstream. Ganapatrao Kulkarni reported that the LTP test cpuset01 in stress mode triggers OOM killer in few seconds, despite lots of free memory. The test attempts to repeatedly fault in memory in one process in a cpuset, while changing allowed nodes of the cpuset between 0 and 1 in another process. One possible cause is that in the fast path we find the preferred zoneref according to current mems_allowed, so that it points to the middle of the zonelist, skipping e.g. zones of node 1 completely. If the mems_allowed is updated to contain only node 1, we never reach it in the zonelist, and trigger OOM before checking the cpuset_mems_cookie. This patch fixes the particular case by redoing the preferred zoneref search if we switch back to the original nodemask. The condition is also slightly changed so that when the last non-root cpuset is removed, we don't miss it. Note that this is not a full fix, and more patches will follow. Link: http://lkml.kernel.org/r/20170120103843.24587-3-vbabka@suse.cz Fixes: 682a3385e773 ("mm, page_alloc: inline the fast path of the zonelist iterator") Signed-off-by: Vlastimil Babka Reported-by: Ganapatrao Kulkarni Acked-by: Michal Hocko Acked-by: Mel Gorman Acked-by: Hillf Danton Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- mm/page_alloc.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 593a11d8bc6b..dedadb4a779f 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3783,9 +3783,17 @@ retry_cpuset: /* * Restore the original nodemask if it was potentially replaced with * &cpuset_current_mems_allowed to optimize the fast-path attempt. + * Also recalculate the starting point for the zonelist iterator or + * we could end up iterating over non-eligible zones endlessly. */ - if (cpusets_enabled()) + if (unlikely(ac.nodemask != nodemask)) { ac.nodemask = nodemask; + ac.preferred_zoneref = first_zones_zonelist(ac.zonelist, + ac.high_zoneidx, ac.nodemask); + if (!ac.preferred_zoneref->zone) + goto no_zone; + } + page = __alloc_pages_slowpath(alloc_mask, order, &ac); no_zone: From b678e4ff7ce0d01bb14f0adb92c1786b0a341cca Mon Sep 17 00:00:00 2001 From: Vlastimil Babka Date: Tue, 24 Jan 2017 15:18:38 -0800 Subject: [PATCH 016/119] mm, page_alloc: move cpuset seqcount checking to slowpath commit 5ce9bfef1d27944c119a397a9d827bef795487ce upstream. This is a preparation for the following patch to make review simpler. While the primary motivation is a bug fix, this also simplifies the fast path, although the moved code is only enabled when cpusets are in use. Link: http://lkml.kernel.org/r/20170120103843.24587-4-vbabka@suse.cz Signed-off-by: Vlastimil Babka Acked-by: Mel Gorman Acked-by: Hillf Danton Cc: Ganapatrao Kulkarni Cc: Michal Hocko Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- mm/page_alloc.c | 47 ++++++++++++++++++++++++++--------------------- 1 file changed, 26 insertions(+), 21 deletions(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index dedadb4a779f..3e04bb398dc6 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3502,12 +3502,13 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order, struct page *page = NULL; unsigned int alloc_flags; unsigned long did_some_progress; - enum compact_priority compact_priority = DEF_COMPACT_PRIORITY; + enum compact_priority compact_priority; enum compact_result compact_result; - int compaction_retries = 0; - int no_progress_loops = 0; + int compaction_retries; + int no_progress_loops; unsigned long alloc_start = jiffies; unsigned int stall_timeout = 10 * HZ; + unsigned int cpuset_mems_cookie; /* * In the slowpath, we sanity check order to avoid ever trying to @@ -3528,6 +3529,12 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order, (__GFP_ATOMIC|__GFP_DIRECT_RECLAIM))) gfp_mask &= ~__GFP_ATOMIC; +retry_cpuset: + compaction_retries = 0; + no_progress_loops = 0; + compact_priority = DEF_COMPACT_PRIORITY; + cpuset_mems_cookie = read_mems_allowed_begin(); + /* * The fast path uses conservative alloc_flags to succeed only until * kswapd needs to be woken up, and to avoid the cost of setting up @@ -3699,6 +3706,15 @@ retry: } nopage: + /* + * When updating a task's mems_allowed, it is possible to race with + * parallel threads in such a way that an allocation can fail while + * the mask is being updated. If a page allocation is about to fail, + * check if the cpuset changed during allocation and if so, retry. + */ + if (read_mems_allowed_retry(cpuset_mems_cookie)) + goto retry_cpuset; + warn_alloc(gfp_mask, "page allocation failure: order:%u", order); got_pg: @@ -3713,7 +3729,6 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist, nodemask_t *nodemask) { struct page *page; - unsigned int cpuset_mems_cookie; unsigned int alloc_flags = ALLOC_WMARK_LOW; gfp_t alloc_mask = gfp_mask; /* The gfp_t that was actually used for allocation */ struct alloc_context ac = { @@ -3750,9 +3765,6 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, if (IS_ENABLED(CONFIG_CMA) && ac.migratetype == MIGRATE_MOVABLE) alloc_flags |= ALLOC_CMA; -retry_cpuset: - cpuset_mems_cookie = read_mems_allowed_begin(); - /* Dirty zone balancing only done in the fast path */ ac.spread_dirty_pages = (gfp_mask & __GFP_WRITE); @@ -3765,6 +3777,11 @@ retry_cpuset: ac.high_zoneidx, ac.nodemask); if (!ac.preferred_zoneref->zone) { page = NULL; + /* + * This might be due to race with cpuset_current_mems_allowed + * update, so make sure we retry with original nodemask in the + * slow path. + */ goto no_zone; } @@ -3773,6 +3790,7 @@ retry_cpuset: if (likely(page)) goto out; +no_zone: /* * Runtime PM, block IO and its error handling path can deadlock * because I/O on the device might not complete. @@ -3790,24 +3808,11 @@ retry_cpuset: ac.nodemask = nodemask; ac.preferred_zoneref = first_zones_zonelist(ac.zonelist, ac.high_zoneidx, ac.nodemask); - if (!ac.preferred_zoneref->zone) - goto no_zone; + /* If we have NULL preferred zone, slowpath wll handle that */ } page = __alloc_pages_slowpath(alloc_mask, order, &ac); -no_zone: - /* - * When updating a task's mems_allowed, it is possible to race with - * parallel threads in such a way that an allocation can fail while - * the mask is being updated. If a page allocation is about to fail, - * check if the cpuset changed during allocation and if so, retry. - */ - if (unlikely(!page && read_mems_allowed_retry(cpuset_mems_cookie))) { - alloc_mask = gfp_mask; - goto retry_cpuset; - } - out: if (memcg_kmem_enabled() && (gfp_mask & __GFP_ACCOUNT) && page && unlikely(memcg_kmem_charge(page, gfp_mask, order) != 0)) { From 96e5cec10e7a75c931f8993633b3a5cedc99144e Mon Sep 17 00:00:00 2001 From: Vlastimil Babka Date: Tue, 24 Jan 2017 15:18:41 -0800 Subject: [PATCH 017/119] mm, page_alloc: fix premature OOM when racing with cpuset mems update commit e47483bca2cc59a4593b37a270b16ee42b1d9f08 upstream. Ganapatrao Kulkarni reported that the LTP test cpuset01 in stress mode triggers OOM killer in few seconds, despite lots of free memory. The test attempts to repeatedly fault in memory in one process in a cpuset, while changing allowed nodes of the cpuset between 0 and 1 in another process. The problem comes from insufficient protection against cpuset changes, which can cause get_page_from_freelist() to consider all zones as non-eligible due to nodemask and/or current->mems_allowed. This was masked in the past by sufficient retries, but since commit 682a3385e773 ("mm, page_alloc: inline the fast path of the zonelist iterator") we fix the preferred_zoneref once, and don't iterate over the whole zonelist in further attempts, thus the only eligible zones might be placed in the zonelist before our starting point and we always miss them. A previous patch fixed this problem for current->mems_allowed. However, cpuset changes also update the task's mempolicy nodemask. The fix has two parts. We have to repeat the preferred_zoneref search when we detect cpuset update by way of seqcount, and we have to check the seqcount before considering OOM. [akpm@linux-foundation.org: fix typo in comment] Link: http://lkml.kernel.org/r/20170120103843.24587-5-vbabka@suse.cz Fixes: c33d6c06f60f ("mm, page_alloc: avoid looking up the first zone in a zonelist twice") Signed-off-by: Vlastimil Babka Reported-by: Ganapatrao Kulkarni Acked-by: Mel Gorman Acked-by: Hillf Danton Cc: Michal Hocko Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- mm/page_alloc.c | 35 ++++++++++++++++++++++++----------- 1 file changed, 24 insertions(+), 11 deletions(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 3e04bb398dc6..f4a02e240fb6 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3534,6 +3534,17 @@ retry_cpuset: no_progress_loops = 0; compact_priority = DEF_COMPACT_PRIORITY; cpuset_mems_cookie = read_mems_allowed_begin(); + /* + * We need to recalculate the starting point for the zonelist iterator + * because we might have used different nodemask in the fast path, or + * there was a cpuset modification and we are retrying - otherwise we + * could end up iterating over non-eligible zones endlessly. + */ + ac->preferred_zoneref = first_zones_zonelist(ac->zonelist, + ac->high_zoneidx, ac->nodemask); + if (!ac->preferred_zoneref->zone) + goto nopage; + /* * The fast path uses conservative alloc_flags to succeed only until @@ -3694,6 +3705,13 @@ retry: &compaction_retries)) goto retry; + /* + * It's possible we raced with cpuset update so the OOM would be + * premature (see below the nopage: label for full explanation). + */ + if (read_mems_allowed_retry(cpuset_mems_cookie)) + goto retry_cpuset; + /* Reclaim has failed us, start killing things */ page = __alloc_pages_may_oom(gfp_mask, order, ac, &did_some_progress); if (page) @@ -3707,10 +3725,11 @@ retry: nopage: /* - * When updating a task's mems_allowed, it is possible to race with - * parallel threads in such a way that an allocation can fail while - * the mask is being updated. If a page allocation is about to fail, - * check if the cpuset changed during allocation and if so, retry. + * When updating a task's mems_allowed or mempolicy nodemask, it is + * possible to race with parallel threads in such a way that our + * allocation can fail while the mask is being updated. If we are about + * to fail, check if the cpuset changed during allocation and if so, + * retry. */ if (read_mems_allowed_retry(cpuset_mems_cookie)) goto retry_cpuset; @@ -3801,15 +3820,9 @@ no_zone: /* * Restore the original nodemask if it was potentially replaced with * &cpuset_current_mems_allowed to optimize the fast-path attempt. - * Also recalculate the starting point for the zonelist iterator or - * we could end up iterating over non-eligible zones endlessly. */ - if (unlikely(ac.nodemask != nodemask)) { + if (unlikely(ac.nodemask != nodemask)) ac.nodemask = nodemask; - ac.preferred_zoneref = first_zones_zonelist(ac.zonelist, - ac.high_zoneidx, ac.nodemask); - /* If we have NULL preferred zone, slowpath wll handle that */ - } page = __alloc_pages_slowpath(alloc_mask, order, &ac); From 13e39d5930ec5a4e0711a6d88bcea84bfd13d4bb Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Fri, 20 Jan 2017 10:33:32 +0000 Subject: [PATCH 018/119] vring: Force use of DMA API for ARM-based systems with legacy devices commit c7070619f3408d9a0dffbed9149e6f00479cf43b upstream. Booting Linux on an ARM fastmodel containing an SMMU emulation results in an unexpected I/O page fault from the legacy virtio-blk PCI device: [ 1.211721] arm-smmu-v3 2b400000.smmu: event 0x10 received: [ 1.211800] arm-smmu-v3 2b400000.smmu: 0x00000000fffff010 [ 1.211880] arm-smmu-v3 2b400000.smmu: 0x0000020800000000 [ 1.211959] arm-smmu-v3 2b400000.smmu: 0x00000008fa081002 [ 1.212075] arm-smmu-v3 2b400000.smmu: 0x0000000000000000 [ 1.212155] arm-smmu-v3 2b400000.smmu: event 0x10 received: [ 1.212234] arm-smmu-v3 2b400000.smmu: 0x00000000fffff010 [ 1.212314] arm-smmu-v3 2b400000.smmu: 0x0000020800000000 [ 1.212394] arm-smmu-v3 2b400000.smmu: 0x00000008fa081000 [ 1.212471] arm-smmu-v3 2b400000.smmu: 0x0000000000000000 This is because the legacy virtio-blk device is behind an SMMU, so we have consequently swizzled its DMA ops and configured the SMMU to translate accesses. This then requires the vring code to use the DMA API to establish translations, otherwise all transactions will result in fatal faults and termination. Given that ARM-based systems only see an SMMU if one is really present (the topology is all described by firmware tables such as device-tree or IORT), then we can safely use the DMA API for all legacy virtio devices. Modern devices can advertise the prescense of an IOMMU using the VIRTIO_F_IOMMU_PLATFORM feature flag. Cc: Andy Lutomirski Cc: Michael S. Tsirkin Fixes: 876945dbf649 ("arm64: Hook up IOMMU dma_ops") Signed-off-by: Will Deacon Signed-off-by: Michael S. Tsirkin Acked-by: Marc Zyngier Signed-off-by: Greg Kroah-Hartman --- drivers/virtio/virtio_ring.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c index 489bfc61cf30..f1360487a594 100644 --- a/drivers/virtio/virtio_ring.c +++ b/drivers/virtio/virtio_ring.c @@ -159,6 +159,13 @@ static bool vring_use_dma_api(struct virtio_device *vdev) if (xen_domain()) return true; + /* + * On ARM-based machines, the DMA ops will do the right thing, + * so always use them with legacy devices. + */ + if (IS_ENABLED(CONFIG_ARM) || IS_ENABLED(CONFIG_ARM64)) + return !virtio_has_feature(vdev, VIRTIO_F_VERSION_1); + return false; } From c7556867782ee8ae0383a1e99df09222f1f01332 Mon Sep 17 00:00:00 2001 From: Nikolay Borisov Date: Fri, 20 Jan 2017 15:21:35 +0200 Subject: [PATCH 019/119] userns: Make ucounts lock irq-safe commit 880a38547ff08715ce4f1daf9a4bb30c87676e68 upstream. The ucounts_lock is being used to protect various ucounts lifecycle management functionalities. However, those services can also be invoked when a pidns is being freed in an RCU callback (e.g. softirq context). This can lead to deadlocks. There were already efforts trying to prevent similar deadlocks in add7c65ca426 ("pid: fix lockdep deadlock warning due to ucount_lock"), however they just moved the context from hardirq to softrq. Fix this issue once and for all by explictly making the lock disable irqs altogether. Dmitry Vyukov reported: > I've got the following deadlock report while running syzkaller fuzzer > on eec0d3d065bfcdf9cd5f56dd2a36b94d12d32297 of linux-next (on odroid > device if it matters): > > ================================= > [ INFO: inconsistent lock state ] > 4.10.0-rc3-next-20170112-xc2-dirty #6 Not tainted > --------------------------------- > inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage. > swapper/2/0 [HC0[0]:SC1[1]:HE1:SE0] takes: > (ucounts_lock){+.?...}, at: [< inline >] spin_lock > ./include/linux/spinlock.h:302 > (ucounts_lock){+.?...}, at: [] > put_ucounts+0x60/0x138 kernel/ucount.c:162 > {SOFTIRQ-ON-W} state was registered at: > [] mark_lock+0x220/0xb60 kernel/locking/lockdep.c:3054 > [< inline >] mark_irqflags kernel/locking/lockdep.c:2941 > [] __lock_acquire+0x388/0x3260 kernel/locking/lockdep.c:3295 > [] lock_acquire+0xa4/0x138 kernel/locking/lockdep.c:3753 > [< inline >] __raw_spin_lock ./include/linux/spinlock_api_smp.h:144 > [] _raw_spin_lock+0x90/0xd0 kernel/locking/spinlock.c:151 > [< inline >] spin_lock ./include/linux/spinlock.h:302 > [< inline >] get_ucounts kernel/ucount.c:131 > [] inc_ucount+0x80/0x6c8 kernel/ucount.c:189 > [< inline >] inc_mnt_namespaces fs/namespace.c:2818 > [] alloc_mnt_ns+0x78/0x3a8 fs/namespace.c:2849 > [] create_mnt_ns+0x28/0x200 fs/namespace.c:2959 > [< inline >] init_mount_tree fs/namespace.c:3199 > [] mnt_init+0x258/0x384 fs/namespace.c:3251 > [] vfs_caches_init+0x6c/0x80 fs/dcache.c:3626 > [] start_kernel+0x414/0x460 init/main.c:648 > [] __primary_switched+0x6c/0x70 arch/arm64/kernel/head.S:456 > irq event stamp: 2316924 > hardirqs last enabled at (2316924): [< inline >] rcu_do_batch > kernel/rcu/tree.c:2911 > hardirqs last enabled at (2316924): [< inline >] > invoke_rcu_callbacks kernel/rcu/tree.c:3182 > hardirqs last enabled at (2316924): [< inline >] > __rcu_process_callbacks kernel/rcu/tree.c:3149 > hardirqs last enabled at (2316924): [] > rcu_process_callbacks+0x7a4/0xc28 kernel/rcu/tree.c:3166 > hardirqs last disabled at (2316923): [< inline >] rcu_do_batch > kernel/rcu/tree.c:2900 > hardirqs last disabled at (2316923): [< inline >] > invoke_rcu_callbacks kernel/rcu/tree.c:3182 > hardirqs last disabled at (2316923): [< inline >] > __rcu_process_callbacks kernel/rcu/tree.c:3149 > hardirqs last disabled at (2316923): [] > rcu_process_callbacks+0x210/0xc28 kernel/rcu/tree.c:3166 > softirqs last enabled at (2316912): [] > _local_bh_enable+0x4c/0x80 kernel/softirq.c:155 > softirqs last disabled at (2316913): [< inline >] > do_softirq_own_stack ./include/linux/interrupt.h:488 > softirqs last disabled at (2316913): [< inline >] > invoke_softirq kernel/softirq.c:371 > softirqs last disabled at (2316913): [] > irq_exit+0x264/0x308 kernel/softirq.c:405 > > other info that might help us debug this: > Possible unsafe locking scenario: > > CPU0 > ---- > lock(ucounts_lock); > > lock(ucounts_lock); > > *** DEADLOCK *** > > 1 lock held by swapper/2/0: > #0: (rcu_callback){......}, at: [< inline >] __rcu_reclaim > kernel/rcu/rcu.h:108 > #0: (rcu_callback){......}, at: [< inline >] rcu_do_batch > kernel/rcu/tree.c:2919 > #0: (rcu_callback){......}, at: [< inline >] > invoke_rcu_callbacks kernel/rcu/tree.c:3182 > #0: (rcu_callback){......}, at: [< inline >] > __rcu_process_callbacks kernel/rcu/tree.c:3149 > #0: (rcu_callback){......}, at: [] > rcu_process_callbacks+0x720/0xc28 kernel/rcu/tree.c:3166 > > stack backtrace: > CPU: 2 PID: 0 Comm: swapper/2 Not tainted 4.10.0-rc3-next-20170112-xc2-dirty #6 > Hardware name: Hardkernel ODROID-C2 (DT) > Call trace: > [] dump_backtrace+0x0/0x440 arch/arm64/kernel/traps.c:500 > [] show_stack+0x20/0x30 arch/arm64/kernel/traps.c:225 > [] dump_stack+0x110/0x168 > [] print_usage_bug.part.27+0x49c/0x4bc > kernel/locking/lockdep.c:2387 > [< inline >] print_usage_bug kernel/locking/lockdep.c:2357 > [< inline >] valid_state kernel/locking/lockdep.c:2400 > [< inline >] mark_lock_irq kernel/locking/lockdep.c:2617 > [] mark_lock+0x934/0xb60 kernel/locking/lockdep.c:3065 > [< inline >] mark_irqflags kernel/locking/lockdep.c:2923 > [] __lock_acquire+0x640/0x3260 kernel/locking/lockdep.c:3295 > [] lock_acquire+0xa4/0x138 kernel/locking/lockdep.c:3753 > [< inline >] __raw_spin_lock ./include/linux/spinlock_api_smp.h:144 > [] _raw_spin_lock+0x90/0xd0 kernel/locking/spinlock.c:151 > [< inline >] spin_lock ./include/linux/spinlock.h:302 > [] put_ucounts+0x60/0x138 kernel/ucount.c:162 > [] dec_ucount+0xf4/0x158 kernel/ucount.c:214 > [< inline >] dec_pid_namespaces kernel/pid_namespace.c:89 > [] delayed_free_pidns+0x40/0xe0 kernel/pid_namespace.c:156 > [< inline >] __rcu_reclaim kernel/rcu/rcu.h:118 > [< inline >] rcu_do_batch kernel/rcu/tree.c:2919 > [< inline >] invoke_rcu_callbacks kernel/rcu/tree.c:3182 > [< inline >] __rcu_process_callbacks kernel/rcu/tree.c:3149 > [] rcu_process_callbacks+0x768/0xc28 kernel/rcu/tree.c:3166 > [] __do_softirq+0x324/0x6e0 kernel/softirq.c:284 > [< inline >] do_softirq_own_stack ./include/linux/interrupt.h:488 > [< inline >] invoke_softirq kernel/softirq.c:371 > [] irq_exit+0x264/0x308 kernel/softirq.c:405 > [] __handle_domain_irq+0xc0/0x150 kernel/irq/irqdesc.c:636 > [] gic_handle_irq+0x68/0xd8 > Exception stack(0xffff8000648e7dd0 to 0xffff8000648e7f00) > 7dc0: ffff8000648d4b3c 0000000000000007 > 7de0: 0000000000000000 1ffff0000c91a967 1ffff0000c91a967 1ffff0000c91a967 > 7e00: ffff20000a4b6b68 0000000000000001 0000000000000007 0000000000000001 > 7e20: 1fffe4000149ae90 ffff200009d35000 0000000000000000 0000000000000002 > 7e40: 0000000000000000 0000000000000000 0000000002624a1a 0000000000000000 > 7e60: 0000000000000000 ffff200009cbcd88 000060006d2ed000 0000000000000140 > 7e80: ffff200009cff000 ffff200009cb6000 ffff200009cc2020 ffff200009d2159d > 7ea0: 0000000000000000 ffff8000648d4380 0000000000000000 ffff8000648e7f00 > 7ec0: ffff20000820a478 ffff8000648e7f00 ffff20000820a47c 0000000010000145 > 7ee0: 0000000000000140 dfff200000000000 ffffffffffffffff ffff20000820a478 > [] el1_irq+0xb8/0x130 arch/arm64/kernel/entry.S:486 > [< inline >] arch_local_irq_restore > ./arch/arm64/include/asm/irqflags.h:81 > [] rcu_idle_exit+0x64/0xa8 kernel/rcu/tree.c:1030 > [< inline >] cpuidle_idle_call kernel/sched/idle.c:200 > [] do_idle+0x1dc/0x2d0 kernel/sched/idle.c:243 > [] cpu_startup_entry+0x24/0x28 kernel/sched/idle.c:345 > [] secondary_start_kernel+0x2cc/0x358 > arch/arm64/kernel/smp.c:276 > [<000000000279f1a4>] 0x279f1a4 Reported-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Fixes: add7c65ca426 ("pid: fix lockdep deadlock warning due to ucount_lock") Fixes: f333c700c610 ("pidns: Add a limit on the number of pid namespaces") Link: https://www.spinics.net/lists/kernel/msg2426637.html Signed-off-by: Nikolay Borisov Signed-off-by: Eric W. Biederman Signed-off-by: Greg Kroah-Hartman --- kernel/ucount.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/kernel/ucount.c b/kernel/ucount.c index 9d20d5dd298a..4bbd38ec3788 100644 --- a/kernel/ucount.c +++ b/kernel/ucount.c @@ -128,10 +128,10 @@ static struct ucounts *get_ucounts(struct user_namespace *ns, kuid_t uid) struct hlist_head *hashent = ucounts_hashentry(ns, uid); struct ucounts *ucounts, *new; - spin_lock(&ucounts_lock); + spin_lock_irq(&ucounts_lock); ucounts = find_ucounts(ns, uid, hashent); if (!ucounts) { - spin_unlock(&ucounts_lock); + spin_unlock_irq(&ucounts_lock); new = kzalloc(sizeof(*new), GFP_KERNEL); if (!new) @@ -141,7 +141,7 @@ static struct ucounts *get_ucounts(struct user_namespace *ns, kuid_t uid) new->uid = uid; atomic_set(&new->count, 0); - spin_lock(&ucounts_lock); + spin_lock_irq(&ucounts_lock); ucounts = find_ucounts(ns, uid, hashent); if (ucounts) { kfree(new); @@ -152,16 +152,18 @@ static struct ucounts *get_ucounts(struct user_namespace *ns, kuid_t uid) } if (!atomic_add_unless(&ucounts->count, 1, INT_MAX)) ucounts = NULL; - spin_unlock(&ucounts_lock); + spin_unlock_irq(&ucounts_lock); return ucounts; } static void put_ucounts(struct ucounts *ucounts) { + unsigned long flags; + if (atomic_dec_and_test(&ucounts->count)) { - spin_lock(&ucounts_lock); + spin_lock_irqsave(&ucounts_lock, flags); hlist_del_init(&ucounts->node); - spin_unlock(&ucounts_lock); + spin_unlock_irqrestore(&ucounts_lock, flags); kfree(ucounts); } From 03707d6c36f9c7355c22e43039c09804e96a02b6 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 25 Jan 2017 18:20:55 -0800 Subject: [PATCH 020/119] sysctl: fix proc_doulongvec_ms_jiffies_minmax() commit ff9f8a7cf935468a94d9927c68b00daae701667e upstream. We perform the conversion between kernel jiffies and ms only when exporting kernel value to user space. We need to do the opposite operation when value is written by user. Only matters when HZ != 1000 Signed-off-by: Eric Dumazet Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- kernel/sysctl.c | 1 + 1 file changed, 1 insertion(+) diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 706309f9ed84..c1095cdc0fe2 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -2487,6 +2487,7 @@ static int __do_proc_doulongvec_minmax(void *data, struct ctl_table *table, int break; if (neg) continue; + val = convmul * val / convdiv; if ((min && val < *min) || (max && val > *max)) continue; *i = val; From 4859524143609fee04d69a8069f9a83317e4bb7c Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Thu, 26 Jan 2017 13:18:09 -0800 Subject: [PATCH 021/119] xfs: prevent quotacheck from overloading inode lru commit e0d76fa4475ef2cf4b52d18588b8ce95153d021b upstream. Quotacheck runs at mount time in situations where quota accounting must be recalculated. In doing so, it uses bulkstat to visit every inode in the filesystem. Historically, every inode processed during quotacheck was released and immediately tagged for reclaim because quotacheck runs before the superblock is marked active by the VFS. In other words, the final iput() lead to an immediate ->destroy_inode() call, which allowed the XFS background reclaim worker to start reclaiming inodes. Commit 17c12bcd3 ("xfs: when replaying bmap operations, don't let unlinked inodes get reaped") marks the XFS superblock active sooner as part of the mount process to support caching inodes processed during log recovery. This occurs before quotacheck and thus means all inodes processed by quotacheck are inserted to the LRU on release. The s_umount lock is held until the mount has completed and thus prevents the shrinkers from operating on the sb. This means that quotacheck can excessively populate the inode LRU and lead to OOM conditions on systems without sufficient RAM. Update the quotacheck bulkstat handler to set XFS_IGET_DONTCACHE on inodes processed by quotacheck. This causes ->drop_inode() to return 1 and in turn causes iput_final() to evict the inode. This preserves the original quotacheck behavior and prevents it from overloading the LRU and running out of memory. Reported-by: Martin Svec Signed-off-by: Brian Foster Reviewed-by: Eric Sandeen Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/xfs_qm.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 45e50ea90769..b669b123287b 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -1177,7 +1177,8 @@ xfs_qm_dqusage_adjust( * the case in all other instances. It's OK that we do this because * quotacheck is done only at mount time. */ - error = xfs_iget(mp, NULL, ino, 0, XFS_ILOCK_EXCL, &ip); + error = xfs_iget(mp, NULL, ino, XFS_IGET_DONTCACHE, XFS_ILOCK_EXCL, + &ip); if (error) { *res = BULKSTAT_RV_NOTHING; return error; From 959f9709c0251275f05fd9092ff25e9b9bec237e Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 27 Jan 2017 13:32:14 +0100 Subject: [PATCH 022/119] ISDN: eicon: silence misleading array-bounds warning commit 950eabbd6ddedc1b08350b9169a6a51b130ebaaf upstream. With some gcc versions, we get a warning about the eicon driver, and that currently shows up as the only remaining warning in one of the build bots: In file included from ../drivers/isdn/hardware/eicon/message.c:30:0: eicon/message.c: In function 'mixer_notify_update': eicon/platform.h:333:18: warning: array subscript is above array bounds [-Warray-bounds] The code is easily changed to open-code the unusual PUT_WORD() line causing this to avoid the warning. Link: http://arm-soc.lixom.net/buildlogs/stable-rc/v4.4.45/ Signed-off-by: Arnd Bergmann Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/isdn/hardware/eicon/message.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/isdn/hardware/eicon/message.c b/drivers/isdn/hardware/eicon/message.c index 1a1d99704fe6..296f1411fe84 100644 --- a/drivers/isdn/hardware/eicon/message.c +++ b/drivers/isdn/hardware/eicon/message.c @@ -11297,7 +11297,8 @@ static void mixer_notify_update(PLCI *plci, byte others) ((CAPI_MSG *) msg)->header.ncci = 0; ((CAPI_MSG *) msg)->info.facility_req.Selector = SELECTOR_LINE_INTERCONNECT; ((CAPI_MSG *) msg)->info.facility_req.structs[0] = 3; - PUT_WORD(&(((CAPI_MSG *) msg)->info.facility_req.structs[1]), LI_REQ_SILENT_UPDATE); + ((CAPI_MSG *) msg)->info.facility_req.structs[1] = LI_REQ_SILENT_UPDATE & 0xff; + ((CAPI_MSG *) msg)->info.facility_req.structs[2] = LI_REQ_SILENT_UPDATE >> 8; ((CAPI_MSG *) msg)->info.facility_req.structs[3] = 0; w = api_put(notify_plci->appl, (CAPI_MSG *) msg); if (w != _QUEUE_FULL) From 79babd4a6ce26d6b0e6a56da47efb723431abc70 Mon Sep 17 00:00:00 2001 From: Omar Sandoval Date: Wed, 25 Jan 2017 17:06:38 -0800 Subject: [PATCH 023/119] Btrfs: remove old tree_root case in btrfs_read_locked_inode() commit 67ade058ef2c65a3e56878af9c293ec76722a2e5 upstream. As Jeff explained in c2951f32d36c ("btrfs: remove old tree_root dirent processing in btrfs_real_readdir()"), supporting this old format is no longer necessary since the Btrfs magic number has been updated since we changed to the current format. There are other places where we still handle this old format, but since this is part of a fix that is going to stable, I'm only removing this one for now. Signed-off-by: Omar Sandoval Reviewed-by: David Sterba Signed-off-by: Chris Mason Signed-off-by: Greg Kroah-Hartman --- fs/btrfs/inode.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 8e3a5a266917..e70b7c695592 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -3819,10 +3819,7 @@ cache_acl: break; case S_IFDIR: inode->i_fop = &btrfs_dir_file_operations; - if (root == root->fs_info->tree_root) - inode->i_op = &btrfs_dir_ro_inode_operations; - else - inode->i_op = &btrfs_dir_inode_operations; + inode->i_op = &btrfs_dir_inode_operations; break; case S_IFLNK: inode->i_op = &btrfs_symlink_inode_operations; From ad80fada9d6d8177d1593a9b5772e80a758db595 Mon Sep 17 00:00:00 2001 From: Omar Sandoval Date: Wed, 25 Jan 2017 17:06:39 -0800 Subject: [PATCH 024/119] Btrfs: disable xattr operations on subvolume directories commit 1fdf41941b8010691679638f8d0c8d08cfee7726 upstream. When you snapshot a subvolume containing a subvolume, you get a placeholder directory where the subvolume would be. These directory inodes have ->i_ops set to btrfs_dir_ro_inode_operations. Previously, these i_ops didn't include the xattr operation callbacks. The conversion to xattr_handlers missed this case, leading to bogus attempts to set xattrs on these inodes. This manifested itself as failures when running delayed inodes. To fix this, clear IOP_XATTR in ->i_opflags on these inodes. Fixes: 6c6ef9f26e59 ("xattr: Stop calling {get,set,remove}xattr inode operations") Cc: Andreas Gruenbacher Reported-by: Chris Murphy Tested-by: Chris Murphy Signed-off-by: Omar Sandoval Reviewed-by: David Sterba Signed-off-by: Chris Mason Signed-off-by: Greg Kroah-Hartman --- fs/btrfs/inode.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index e70b7c695592..68fed83dc742 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5679,6 +5679,7 @@ static struct inode *new_simple_dir(struct super_block *s, inode->i_ino = BTRFS_EMPTY_SUBVOL_DIR_OBJECTID; inode->i_op = &btrfs_dir_ro_inode_operations; + inode->i_opflags &= ~IOP_XATTR; inode->i_fop = &simple_dir_operations; inode->i_mode = S_IFDIR | S_IRUGO | S_IWUSR | S_IXUGO; inode->i_mtime = current_time(inode); From ffb97c11d05fea9a3ff29ad1d9e9c854e0a06dc2 Mon Sep 17 00:00:00 2001 From: Omar Sandoval Date: Wed, 25 Jan 2017 17:06:40 -0800 Subject: [PATCH 025/119] Btrfs: remove ->{get, set}_acl() from btrfs_dir_ro_inode_operations commit 57b59ed2e5b91e958843609c7884794e29e6c4cb upstream. Subvolume directory inodes can't have ACLs. Signed-off-by: Omar Sandoval Reviewed-by: David Sterba Signed-off-by: Chris Mason Signed-off-by: Greg Kroah-Hartman --- fs/btrfs/inode.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 68fed83dc742..be4da91d880f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -10585,8 +10585,6 @@ static const struct inode_operations btrfs_dir_inode_operations = { static const struct inode_operations btrfs_dir_ro_inode_operations = { .lookup = btrfs_lookup, .permission = btrfs_permission, - .get_acl = btrfs_get_acl, - .set_acl = btrfs_set_acl, .update_time = btrfs_update_time, }; From 97a2e39b7ab93008c8bc432675ee36ad3de3fce1 Mon Sep 17 00:00:00 2001 From: Jack Morgenstein Date: Sun, 15 Jan 2017 20:15:00 +0200 Subject: [PATCH 026/119] RDMA/cma: Fix unknown symbol when CONFIG_IPV6 is not enabled commit b4cfe3971f6eab542dd7ecc398bfa1aeec889934 upstream. If IPV6 has not been enabled in the underlying kernel, we must avoid calling IPV6 procedures in rdma_cm.ko. This requires using "IS_ENABLED(CONFIG_IPV6)" in "if" statements surrounding any code which calls external IPV6 procedures. In the instance fixed here, procedure cma_bind_addr() called ipv6_addr_type() -- which resulted in calling external procedure __ipv6_addr_type(). Fixes: 6c26a77124ff ("RDMA/cma: fix IPv6 address resolution") Cc: Spencer Baugh Signed-off-by: Jack Morgenstein Reviewed-by: Moni Shoua Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/core/cma.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 2a6fc47a1dfb..c25768c2dd3b 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -2768,7 +2768,8 @@ static int cma_bind_addr(struct rdma_cm_id *id, struct sockaddr *src_addr, if (!src_addr || !src_addr->sa_family) { src_addr = (struct sockaddr *) &id->route.addr.src_addr; src_addr->sa_family = dst_addr->sa_family; - if (dst_addr->sa_family == AF_INET6) { + if (IS_ENABLED(CONFIG_IPV6) && + dst_addr->sa_family == AF_INET6) { struct sockaddr_in6 *src_addr6 = (struct sockaddr_in6 *) src_addr; struct sockaddr_in6 *dst_addr6 = (struct sockaddr_in6 *) dst_addr; src_addr6->sin6_scope_id = dst_addr6->sin6_scope_id; From 62d7f2123f19a7e81c53bcd08f2902cd54c52b64 Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Mon, 23 Jan 2017 22:59:44 +0100 Subject: [PATCH 027/119] s390/mm: Fix cmma unused transfer from pgste into pte commit 0d6da872d3e4a60f43c295386d7ff9a4cdcd57e9 upstream. The last pgtable rework silently disabled the CMMA unused state by setting a local pte variable (a parameter) instead of propagating it back into the caller. Fix it. Fixes: ebde765c0e85 ("s390/mm: uninline ptep_xxx functions from pgtable.h") Cc: Martin Schwidefsky Cc: Claudio Imbrenda Signed-off-by: Christian Borntraeger Signed-off-by: Martin Schwidefsky Signed-off-by: Greg Kroah-Hartman --- arch/s390/mm/pgtable.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/s390/mm/pgtable.c b/arch/s390/mm/pgtable.c index 7a1897c51c54..d56ef26d4681 100644 --- a/arch/s390/mm/pgtable.c +++ b/arch/s390/mm/pgtable.c @@ -202,7 +202,7 @@ static inline pgste_t ptep_xchg_start(struct mm_struct *mm, return pgste; } -static inline void ptep_xchg_commit(struct mm_struct *mm, +static inline pte_t ptep_xchg_commit(struct mm_struct *mm, unsigned long addr, pte_t *ptep, pgste_t pgste, pte_t old, pte_t new) { @@ -220,6 +220,7 @@ static inline void ptep_xchg_commit(struct mm_struct *mm, } else { *ptep = new; } + return old; } pte_t ptep_xchg_direct(struct mm_struct *mm, unsigned long addr, @@ -231,7 +232,7 @@ pte_t ptep_xchg_direct(struct mm_struct *mm, unsigned long addr, preempt_disable(); pgste = ptep_xchg_start(mm, addr, ptep); old = ptep_flush_direct(mm, addr, ptep); - ptep_xchg_commit(mm, addr, ptep, pgste, old, new); + old = ptep_xchg_commit(mm, addr, ptep, pgste, old, new); preempt_enable(); return old; } @@ -246,7 +247,7 @@ pte_t ptep_xchg_lazy(struct mm_struct *mm, unsigned long addr, preempt_disable(); pgste = ptep_xchg_start(mm, addr, ptep); old = ptep_flush_lazy(mm, addr, ptep); - ptep_xchg_commit(mm, addr, ptep, pgste, old, new); + old = ptep_xchg_commit(mm, addr, ptep, pgste, old, new); preempt_enable(); return old; } From d7f56ee1198a1d0657368abc4dcd10ae109a8433 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Tue, 24 Jan 2017 08:05:52 +0100 Subject: [PATCH 028/119] s390/ptrace: Preserve previous registers for short regset write commit 9dce990d2cf57b5ed4e71a9cdbd7eae4335111ff upstream. Ensure that if userspace supplies insufficient data to PTRACE_SETREGSET to fill all the registers, the thread's old registers are preserved. convert_vx_to_fp() is adapted to handle only a specified number of registers rather than unconditionally handling all of them: other callers of this function are adapted appropriately. Based on an initial patch by Dave Martin. Reported-by: Dave Martin Signed-off-by: Martin Schwidefsky Signed-off-by: Greg Kroah-Hartman --- arch/s390/kernel/ptrace.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c index 9336e824e2db..fc2974b929c3 100644 --- a/arch/s390/kernel/ptrace.c +++ b/arch/s390/kernel/ptrace.c @@ -963,6 +963,11 @@ static int s390_fpregs_set(struct task_struct *target, if (target == current) save_fpu_regs(); + if (MACHINE_HAS_VX) + convert_vx_to_fp(fprs, target->thread.fpu.vxrs); + else + memcpy(&fprs, target->thread.fpu.fprs, sizeof(fprs)); + /* If setting FPC, must validate it first. */ if (count > 0 && pos < offsetof(s390_fp_regs, fprs)) { u32 ufpc[2] = { target->thread.fpu.fpc, 0 }; @@ -1067,6 +1072,9 @@ static int s390_vxrs_low_set(struct task_struct *target, if (target == current) save_fpu_regs(); + for (i = 0; i < __NUM_VXRS_LOW; i++) + vxrs[i] = *((__u64 *)(target->thread.fpu.vxrs + i) + 1); + rc = user_regset_copyin(&pos, &count, &kbuf, &ubuf, vxrs, 0, -1); if (rc == 0) for (i = 0; i < __NUM_VXRS_LOW; i++) From 95600605ffaa644d3f6fce55faf40ec12cec7855 Mon Sep 17 00:00:00 2001 From: Nicolas Iooss Date: Sun, 22 Jan 2017 14:41:22 +0100 Subject: [PATCH 029/119] IB/cxgb3: fix misspelling in header guard commit b1a27eac7fefff33ccf6acc919fc0725bf9815fb upstream. Use CXGB3_... instead of CXBG3_... Fixes: a85fb3383340 ("IB/cxgb3: Move user vendor structures") Signed-off-by: Nicolas Iooss Reviewed-by: Leon Romanovsky Acked-by: Steve Wise Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- include/uapi/rdma/cxgb3-abi.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/uapi/rdma/cxgb3-abi.h b/include/uapi/rdma/cxgb3-abi.h index 48a19bda071b..d24eee12128f 100644 --- a/include/uapi/rdma/cxgb3-abi.h +++ b/include/uapi/rdma/cxgb3-abi.h @@ -30,7 +30,7 @@ * SOFTWARE. */ #ifndef CXGB3_ABI_USER_H -#define CXBG3_ABI_USER_H +#define CXGB3_ABI_USER_H #include From 24be606cd3b433412488ce962842489fe6e34a7a Mon Sep 17 00:00:00 2001 From: Max Gurtovoy Date: Wed, 18 Jan 2017 00:40:39 +0200 Subject: [PATCH 030/119] IB/iser: Fix sg_tablesize calculation commit 1e5db6c31ade4150c2e2b1a21e39f776c38fea39 upstream. For devices that can register page list that is bigger than USHRT_MAX, we actually take the wrong value for sg_tablesize. E.g: for CX4 max_fast_reg_page_list_len is 65536 (bigger than USHRT_MAX) so we set sg_tablesize to 0 by mistake. Therefore, each IO that is bigger than 4k splitted to "< 4k" chunks that cause performance degredation. Remove wrong sg_tablesize assignment, and use the value that was set during address resolution handler with the needed casting. Signed-off-by: Max Gurtovoy Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/ulp/iser/iscsi_iser.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.c b/drivers/infiniband/ulp/iser/iscsi_iser.c index 64b3d11dcf1e..140f3f354cf3 100644 --- a/drivers/infiniband/ulp/iser/iscsi_iser.c +++ b/drivers/infiniband/ulp/iser/iscsi_iser.c @@ -651,13 +651,6 @@ iscsi_iser_session_create(struct iscsi_endpoint *ep, SHOST_DIX_GUARD_CRC); } - /* - * Limit the sg_tablesize and max_sectors based on the device - * max fastreg page list length. - */ - shost->sg_tablesize = min_t(unsigned short, shost->sg_tablesize, - ib_conn->device->ib_device->attrs.max_fast_reg_page_list_len); - if (iscsi_host_add(shost, ib_conn->device->ib_device->dma_device)) { mutex_unlock(&iser_conn->state_mutex); From c2293e76babeebea894f584c11a07290b4f701da Mon Sep 17 00:00:00 2001 From: Israel Rukshin Date: Wed, 28 Dec 2016 12:48:28 +0200 Subject: [PATCH 031/119] IB/srp: fix mr allocation when the device supports sg gaps commit ad8e66b4a80182174f73487ed25fd2140cf43361 upstream. If the device support arbitrary sg list mapping (device cap IB_DEVICE_SG_GAPS_REG set) we allocate the memory regions with IB_MR_TYPE_SG_GAPS. Fixes: 509c5f33f4f6 ("IB/srp: Prevent mapping failures") Signed-off-by: Israel Rukshin Signed-off-by: Max Gurtovoy Reviewed-by: Leon Romanovsky Reviewed-by: Mark Bloch Reviewed-by: Yuval Shaia Reviewed-by: Bart Van Assche Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/ulp/srp/ib_srp.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index d980fb458ad4..69d2d6fb73c1 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -366,6 +366,7 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, struct srp_fr_desc *d; struct ib_mr *mr; int i, ret = -EINVAL; + enum ib_mr_type mr_type; if (pool_size <= 0) goto err; @@ -379,9 +380,13 @@ static struct srp_fr_pool *srp_create_fr_pool(struct ib_device *device, spin_lock_init(&pool->lock); INIT_LIST_HEAD(&pool->free_list); + if (device->attrs.device_cap_flags & IB_DEVICE_SG_GAPS_REG) + mr_type = IB_MR_TYPE_SG_GAPS; + else + mr_type = IB_MR_TYPE_MEM_REG; + for (i = 0, d = &pool->desc[0]; i < pool->size; i++, d++) { - mr = ib_alloc_mr(pd, IB_MR_TYPE_MEM_REG, - max_page_list_len); + mr = ib_alloc_mr(pd, mr_type, max_page_list_len); if (IS_ERR(mr)) { ret = PTR_ERR(mr); goto destroy_pool; From a1af471b4151176e26330d3c4a9589711866e2b1 Mon Sep 17 00:00:00 2001 From: Israel Rukshin Date: Wed, 4 Jan 2017 15:59:37 +0200 Subject: [PATCH 032/119] IB/srp: fix invalid indirect_sg_entries parameter value commit 0a475ef4226e305bdcffe12b401ca1eab06c4913 upstream. After setting indirect_sg_entries module_param to huge value (e.g 500,000), srp_alloc_req_data() fails to allocate indirect descriptors for the request ring (kmalloc fails). This commit enforces the maximum value of indirect_sg_entries to be SG_MAX_SEGMENTS as signified in module param description. Fixes: 65e8617fba17 (scsi: rename SCSI_MAX_{SG, SG_CHAIN}_SEGMENTS) Fixes: c07d424d6118 (IB/srp: add support for indirect tables that don't fit in SRP_CMD) Signed-off-by: Israel Rukshin Signed-off-by: Max Gurtovoy Reviewed-by: Laurence Oberman Reviewed-by: Bart Van Assche -- Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/ulp/srp/ib_srp.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 69d2d6fb73c1..e7dcf14a76e2 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -3683,6 +3683,12 @@ static int __init srp_init_module(void) indirect_sg_entries = cmd_sg_entries; } + if (indirect_sg_entries > SG_MAX_SEGMENTS) { + pr_warn("Clamping indirect_sg_entries to %u\n", + SG_MAX_SEGMENTS); + indirect_sg_entries = SG_MAX_SEGMENTS; + } + srp_remove_wq = create_workqueue("srp_remove"); if (!srp_remove_wq) { ret = -ENOMEM; From 9f56548b007769191f7dff7585eb317369991c22 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Einar=20J=C3=B3n?= Date: Fri, 12 Aug 2016 13:50:41 +0200 Subject: [PATCH 033/119] can: c_can_pci: fix null-pointer-deref in c_can_start() - set device pointer MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit c97c52be78b8463ac5407f1cf1f22f8f6cf93a37 upstream. The priv->device pointer for c_can_pci is never set, but it is used without a NULL check in c_can_start(). Setting it in c_can_pci_probe() like c_can_plat_probe() prevents c_can_pci.ko from crashing, with and without CONFIG_PM. This might also cause the pm_runtime_*() functions in c_can.c to actually be executed for c_can_pci devices - they are the only other place where priv->device is used, but they all contain a null check. Signed-off-by: Einar Jón Signed-off-by: Marc Kleine-Budde Signed-off-by: Greg Kroah-Hartman --- drivers/net/can/c_can/c_can_pci.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/can/c_can/c_can_pci.c b/drivers/net/can/c_can/c_can_pci.c index 7be393c96b1a..cf7c18947189 100644 --- a/drivers/net/can/c_can/c_can_pci.c +++ b/drivers/net/can/c_can/c_can_pci.c @@ -161,6 +161,7 @@ static int c_can_pci_probe(struct pci_dev *pdev, dev->irq = pdev->irq; priv->base = addr; + priv->device = &pdev->dev; if (!c_can_pci_data->freq) { dev_err(&pdev->dev, "no clock frequency defined\n"); From 50f5972cc23179e90e8452998bcb609694a9d927 Mon Sep 17 00:00:00 2001 From: Yegor Yefremov Date: Wed, 18 Jan 2017 11:35:57 +0100 Subject: [PATCH 034/119] can: ti_hecc: add missing prepare and unprepare of the clock commit befa60113ce7ea270cb51eada28443ca2756f480 upstream. In order to make the driver work with the common clock framework, this patch converts the clk_enable()/clk_disable() to clk_prepare_enable()/clk_disable_unprepare(). Also add error checking for clk_prepare_enable(). Signed-off-by: Yegor Yefremov Signed-off-by: Marc Kleine-Budde Signed-off-by: Greg Kroah-Hartman --- drivers/net/can/ti_hecc.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/drivers/net/can/ti_hecc.c b/drivers/net/can/ti_hecc.c index 680d1ff07a55..6749b1829469 100644 --- a/drivers/net/can/ti_hecc.c +++ b/drivers/net/can/ti_hecc.c @@ -948,7 +948,12 @@ static int ti_hecc_probe(struct platform_device *pdev) netif_napi_add(ndev, &priv->napi, ti_hecc_rx_poll, HECC_DEF_NAPI_WEIGHT); - clk_enable(priv->clk); + err = clk_prepare_enable(priv->clk); + if (err) { + dev_err(&pdev->dev, "clk_prepare_enable() failed\n"); + goto probe_exit_clk; + } + err = register_candev(ndev); if (err) { dev_err(&pdev->dev, "register_candev() failed\n"); @@ -981,7 +986,7 @@ static int ti_hecc_remove(struct platform_device *pdev) struct ti_hecc_priv *priv = netdev_priv(ndev); unregister_candev(ndev); - clk_disable(priv->clk); + clk_disable_unprepare(priv->clk); clk_put(priv->clk); res = platform_get_resource(pdev, IORESOURCE_MEM, 0); iounmap(priv->base); @@ -1006,7 +1011,7 @@ static int ti_hecc_suspend(struct platform_device *pdev, pm_message_t state) hecc_set_bit(priv, HECC_CANMC, HECC_CANMC_PDR); priv->can.state = CAN_STATE_SLEEPING; - clk_disable(priv->clk); + clk_disable_unprepare(priv->clk); return 0; } @@ -1015,8 +1020,11 @@ static int ti_hecc_resume(struct platform_device *pdev) { struct net_device *dev = platform_get_drvdata(pdev); struct ti_hecc_priv *priv = netdev_priv(dev); + int err; - clk_enable(priv->clk); + err = clk_prepare_enable(priv->clk); + if (err) + return err; hecc_clear_bit(priv, HECC_CANMC, HECC_CANMC_PDR); priv->can.state = CAN_STATE_ERROR_ACTIVE; From 9d5f2c151ec01949e4cc02333bc843b128110f55 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Tue, 24 Jan 2017 10:23:42 -0800 Subject: [PATCH 035/119] ARC: udelay: fix inline assembler by adding LP_COUNT to clobber list commit 36425cd67052e3becf325fd4d3ba5691791ef7e4 upstream. commit 3c7c7a2fc8811bc ("ARC: Don't use "+l" inline asm constraint") modified the inline assembly to setup LP_COUNT register manually and NOT rely on gcc to do it (with the +l inline assembler contraint hint, now being retired in the compiler) However the fix was flawed as we didn't add LP_COUNT to asm clobber list, meaning gcc doesn't know that LP_COUNT or zero-delay-loops are in action in the inline asm. This resulted in some fun - as nested ZOL loops were being generared | mov lp_count,250000 ;16 # tmp235, | lp .L__GCC__LP14 # <======= OUTER LOOP (gcc generated) | .L14: | ld r2, [r5] # MEM[(volatile u32 *)prephitmp_43], w | dmb 1 | breq r2, -1, @.L21 #, w,, | bbit0 r2,1,@.L13 # w,, | ld r4,[r7] ;25 # loops_per_jiffy, loops_per_jiffy | mpymu r3,r4,r6 #, loops_per_jiffy, tmp234 | | mov lp_count, r3 # <====== INNER LOOP (from inline asm) | lp 1f | nop | 1: | nop_s | .L__GCC__LP14: ; loop end, start is @.L14 #, This caused issues with drivers relying on sane behaviour of udelay friends. With LP_COUNT added to clobber list, gcc doesn't generate the outer loop in say above case. Addresses STAR 9001146134 Reported-by: Joao Pinto Fixes: 3c7c7a2fc8811bc ("ARC: Don't use "+l" inline asm constraint") Signed-off-by: Vineet Gupta Signed-off-by: Greg Kroah-Hartman --- arch/arc/include/asm/delay.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/arc/include/asm/delay.h b/arch/arc/include/asm/delay.h index a36e8601114d..d5da2115d78a 100644 --- a/arch/arc/include/asm/delay.h +++ b/arch/arc/include/asm/delay.h @@ -26,7 +26,9 @@ static inline void __delay(unsigned long loops) " lp 1f \n" " nop \n" "1: \n" - : : "r"(loops)); + : + : "r"(loops) + : "lp_count"); } extern void __bad_udelay(void); From ca332b96ba623aa3243a79d355684974f88c3f1a Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Fri, 27 Jan 2017 10:45:27 -0800 Subject: [PATCH 036/119] ARC: [arcompact] handle unaligned access delay slot corner case commit 9aed02feae57bf7a40cb04ea0e3017cb7a998db4 upstream. After emulating an unaligned access in delay slot of a branch, we pretend as the delay slot never happened - so return back to actual branch target (or next PC if branch was not taken). Curently we did this by handling STATUS32.DE, we also need to clear the BTA.T bit, which is disregarded when returning from original misaligned exception, but could cause weirdness if it took the interrupt return path (in case interrupt was acive too) One ARC700 customer ran into this when enabling unaligned access fixup for kernel mode accesses as well Signed-off-by: Vineet Gupta Signed-off-by: Greg Kroah-Hartman --- arch/arc/kernel/unaligned.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arc/kernel/unaligned.c b/arch/arc/kernel/unaligned.c index abd961f3e763..91ebe382147f 100644 --- a/arch/arc/kernel/unaligned.c +++ b/arch/arc/kernel/unaligned.c @@ -241,8 +241,9 @@ int misaligned_fixup(unsigned long address, struct pt_regs *regs, if (state.fault) goto fault; + /* clear any remanants of delay slot */ if (delay_mode(regs)) { - regs->ret = regs->bta; + regs->ret = regs->bta ~1U; regs->status32 &= ~STATUS_DE_MASK; } else { regs->ret += state.instr_len; From 2b95f1210e5010feb767edefcfff53f2c611c1c5 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sat, 28 Jan 2017 11:52:02 +0100 Subject: [PATCH 037/119] parisc: Don't use BITS_PER_LONG in userspace-exported swab.h header commit 2ad5d52d42810bed95100a3d912679d8864421ec upstream. In swab.h the "#if BITS_PER_LONG > 32" breaks compiling userspace programs if BITS_PER_LONG is #defined by userspace with the sizeof() compiler builtin. Solve this problem by using __BITS_PER_LONG instead. Since we now #include asm/bitsperlong.h avoid further potential userspace pollution by moving the #define of SHIFT_PER_LONG to bitops.h which is not exported to userspace. This patch unbreaks compiling qemu on hppa/parisc. Signed-off-by: Helge Deller Signed-off-by: Greg Kroah-Hartman --- arch/parisc/include/asm/bitops.h | 8 +++++++- arch/parisc/include/uapi/asm/bitsperlong.h | 2 -- arch/parisc/include/uapi/asm/swab.h | 5 +++-- 3 files changed, 10 insertions(+), 5 deletions(-) diff --git a/arch/parisc/include/asm/bitops.h b/arch/parisc/include/asm/bitops.h index 3f9406d9b9d6..da87943328a5 100644 --- a/arch/parisc/include/asm/bitops.h +++ b/arch/parisc/include/asm/bitops.h @@ -6,7 +6,7 @@ #endif #include -#include /* for BITS_PER_LONG/SHIFT_PER_LONG */ +#include #include #include #include @@ -17,6 +17,12 @@ * to include/asm-i386/bitops.h or kerneldoc */ +#if __BITS_PER_LONG == 64 +#define SHIFT_PER_LONG 6 +#else +#define SHIFT_PER_LONG 5 +#endif + #define CHOP_SHIFTCOUNT(x) (((unsigned long) (x)) & (BITS_PER_LONG - 1)) diff --git a/arch/parisc/include/uapi/asm/bitsperlong.h b/arch/parisc/include/uapi/asm/bitsperlong.h index e0a23c7bdd43..07fa7e50bdc0 100644 --- a/arch/parisc/include/uapi/asm/bitsperlong.h +++ b/arch/parisc/include/uapi/asm/bitsperlong.h @@ -3,10 +3,8 @@ #if defined(__LP64__) #define __BITS_PER_LONG 64 -#define SHIFT_PER_LONG 6 #else #define __BITS_PER_LONG 32 -#define SHIFT_PER_LONG 5 #endif #include diff --git a/arch/parisc/include/uapi/asm/swab.h b/arch/parisc/include/uapi/asm/swab.h index e78403b129ef..928e1bbac98f 100644 --- a/arch/parisc/include/uapi/asm/swab.h +++ b/arch/parisc/include/uapi/asm/swab.h @@ -1,6 +1,7 @@ #ifndef _PARISC_SWAB_H #define _PARISC_SWAB_H +#include #include #include @@ -38,7 +39,7 @@ static inline __attribute_const__ __u32 __arch_swab32(__u32 x) } #define __arch_swab32 __arch_swab32 -#if BITS_PER_LONG > 32 +#if __BITS_PER_LONG > 32 /* ** From "PA-RISC 2.0 Architecture", HP Professional Books. ** See Appendix I page 8 , "Endian Byte Swapping". @@ -61,6 +62,6 @@ static inline __attribute_const__ __u64 __arch_swab64(__u64 x) return x; } #define __arch_swab64 __arch_swab64 -#endif /* BITS_PER_LONG > 32 */ +#endif /* __BITS_PER_LONG > 32 */ #endif /* _PARISC_SWAB_H */ From 73fdda3b01cd76c6570a4146bfdc6703cfce73ee Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Sun, 22 Jan 2017 14:04:29 -0500 Subject: [PATCH 038/119] nfs: Don't increment lock sequence ID after NFS4ERR_MOVED commit 059aa734824165507c65fd30a55ff000afd14983 upstream. Xuan Qi reports that the Linux NFSv4 client failed to lock a file that was migrated. The steps he observed on the wire: 1. The client sent a LOCK request to the source server 2. The source server replied NFS4ERR_MOVED 3. The client switched to the destination server 4. The client sent the same LOCK request to the destination server with a bumped lock sequence ID 5. The destination server rejected the LOCK request with NFS4ERR_BAD_SEQID RFC 3530 section 8.1.5 provides a list of NFS errors which do not bump a lock sequence ID. However, RFC 3530 is now obsoleted by RFC 7530. In RFC 7530 section 9.1.7, this list has been updated by the addition of NFS4ERR_MOVED. Reported-by: Xuan Qi Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust Signed-off-by: Greg Kroah-Hartman --- include/linux/nfs4.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index 9094faf0699d..039e76e91896 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -282,7 +282,7 @@ enum nfsstat4 { static inline bool seqid_mutating_err(u32 err) { - /* rfc 3530 section 8.1.5: */ + /* See RFC 7530, section 9.1.7 */ switch (err) { case NFS4ERR_STALE_CLIENTID: case NFS4ERR_STALE_STATEID: @@ -291,6 +291,7 @@ static inline bool seqid_mutating_err(u32 err) case NFS4ERR_BADXDR: case NFS4ERR_RESOURCE: case NFS4ERR_NOFILEHANDLE: + case NFS4ERR_MOVED: return false; }; return true; From 0a7023506112ea7c634dffe08683d5d90f52eec6 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Mon, 23 Jan 2017 22:44:12 -0500 Subject: [PATCH 039/119] NFSv4.1: Fix a deadlock in layoutget commit 8ac092519ad91931c96d306c4bfae2c6587c325f upstream. We cannot call nfs4_handle_exception() without first ensuring that the slot has been freed. If not, we end up deadlocking with the process waiting for recovery to complete, and recovery waiting for the slot table to drain. Fixes: 2e80dbe7ac51 ("NFSv4.1: Close callback races for OPEN, LAYOUTGET...") Signed-off-by: Trond Myklebust Signed-off-by: Greg Kroah-Hartman --- fs/nfs/nfs4proc.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 241da19b7da4..e475a3d13f27 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -8371,6 +8371,7 @@ nfs4_layoutget_handle_exception(struct rpc_task *task, goto out; } + nfs4_sequence_free_slot(&lgp->res.seq_res); err = nfs4_handle_exception(server, nfs4err, exception); if (!status) { if (exception->retry) From 5637949edb50c54b3074f76b648d7f873d8a6814 Mon Sep 17 00:00:00 2001 From: Benjamin Coddington Date: Tue, 24 Jan 2017 11:34:20 -0500 Subject: [PATCH 040/119] NFSv4.0: always send mode in SETATTR after EXCLUSIVE4 commit a430607b2ef7c3be090f88c71cfcb1b3988aa7c0 upstream. Some nfsv4.0 servers may return a mode for the verifier following an open with EXCLUSIVE4 createmode, but this does not mean the client should skip setting the mode in the following SETATTR. It should only do that for EXCLUSIVE4_1 or UNGAURDED createmode. Fixes: 5334c5bdac92 ("NFS: Send attributes in OPEN request for NFS4_CREATE_EXCLUSIVE4_1") Signed-off-by: Benjamin Coddington Signed-off-by: Trond Myklebust Signed-off-by: Greg Kroah-Hartman --- fs/nfs/nfs4proc.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index e475a3d13f27..78ff8b63d5f7 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2678,7 +2678,8 @@ static inline void nfs4_exclusive_attrset(struct nfs4_opendata *opendata, sattr->ia_valid |= ATTR_MTIME; /* Except MODE, it seems harmless of setting twice. */ - if ((attrset[1] & FATTR4_WORD1_MODE)) + if (opendata->o_arg.createmode != NFS4_CREATE_EXCLUSIVE && + attrset[1] & FATTR4_WORD1_MODE) sattr->ia_valid &= ~ATTR_MODE; if (attrset[2] & FATTR4_WORD2_SECURITY_LABEL) From cb1d48f55a6dd1ad04caec1c140c2a136eb99206 Mon Sep 17 00:00:00 2001 From: Kinglong Mee Date: Fri, 20 Jan 2017 16:48:39 +0800 Subject: [PATCH 041/119] SUNRPC: cleanup ida information when removing sunrpc module commit c929ea0b910355e1876c64431f3d5802f95b3d75 upstream. After removing sunrpc module, I get many kmemleak information as, unreferenced object 0xffff88003316b1e0 (size 544): comm "gssproxy", pid 2148, jiffies 4294794465 (age 4200.081s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [] kmemleak_alloc+0x4a/0xa0 [] kmem_cache_alloc+0x15e/0x1f0 [] ida_pre_get+0xaa/0x150 [] ida_simple_get+0xad/0x180 [] nlmsvc_lookup_host+0x4ab/0x7f0 [lockd] [] lockd+0x4d/0x270 [lockd] [] param_set_timeout+0x55/0x100 [lockd] [] svc_defer+0x114/0x3f0 [sunrpc] [] svc_defer+0x2d7/0x3f0 [sunrpc] [] rpc_show_info+0x8a/0x110 [sunrpc] [] proc_reg_write+0x7f/0xc0 [] __vfs_write+0xdf/0x3c0 [] vfs_write+0xef/0x240 [] SyS_write+0xad/0x130 [] entry_SYSCALL_64_fastpath+0x1a/0xa9 [] 0xffffffffffffffff I found, the ida information (dynamic memory) isn't cleanup. Signed-off-by: Kinglong Mee Fixes: 2f048db4680a ("SUNRPC: Add an identifier for struct rpc_clnt") Signed-off-by: Trond Myklebust Signed-off-by: Greg Kroah-Hartman --- include/linux/sunrpc/clnt.h | 1 + net/sunrpc/clnt.c | 5 +++++ net/sunrpc/sunrpc_syms.c | 1 + 3 files changed, 7 insertions(+) diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index 85cc819676e8..333ad11b3dd9 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -216,5 +216,6 @@ void rpc_clnt_xprt_switch_put(struct rpc_clnt *); void rpc_clnt_xprt_switch_add_xprt(struct rpc_clnt *, struct rpc_xprt *); bool rpc_clnt_xprt_switch_has_addr(struct rpc_clnt *clnt, const struct sockaddr *sap); +void rpc_cleanup_clids(void); #endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_CLNT_H */ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 62a482790937..b2ae4f150ec6 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -336,6 +336,11 @@ out: static DEFINE_IDA(rpc_clids); +void rpc_cleanup_clids(void) +{ + ida_destroy(&rpc_clids); +} + static int rpc_alloc_clid(struct rpc_clnt *clnt) { int clid; diff --git a/net/sunrpc/sunrpc_syms.c b/net/sunrpc/sunrpc_syms.c index ee5d3d253102..3142f38d1104 100644 --- a/net/sunrpc/sunrpc_syms.c +++ b/net/sunrpc/sunrpc_syms.c @@ -119,6 +119,7 @@ out: static void __exit cleanup_sunrpc(void) { + rpc_cleanup_clids(); rpcauth_remove_module(); cleanup_socket_xprt(); svc_cleanup_xprt_sock(); From 0bd3cb8d470b7fdf0fbf677fcc19da915e740edc Mon Sep 17 00:00:00 2001 From: Steve Wise Date: Thu, 22 Dec 2016 07:40:36 -0800 Subject: [PATCH 042/119] iw_cxgb4: free EQ queue memory on last deref commit c12a67fec8d99bb554e8d4e99120d418f1a39c87 upstream. Commit ad61a4c7a9b7 ("iw_cxgb4: don't block in destroy_qp awaiting the last deref") introduced a bug where the RDMA QP EQ queue memory (and QIDs) are possibly freed before the underlying connection has been fully shutdown. The result being a possible DMA read issued by HW after the queue memory has been unmapped and freed. This results in possible WR corruption in the worst case, system bus errors if an IOMMU is in use, and SGE "bad WR" errors reported in the very least. The fix is to defer unmap/free of queue memory and QID resources until the QP struct has been fully dereferenced. To do this, the c4iw_ucontext must also be kept around until the last QP that references it is fully freed. In addition, since the last QP deref can happen in an IRQ disabled context, we need a new workqueue thread to do the final unmap/free of the EQ queue memory. Fixes: ad61a4c7a9b7 ("iw_cxgb4: don't block in destroy_qp awaiting the last deref") Signed-off-by: Steve Wise Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/hw/cxgb4/device.c | 9 +++++++ drivers/infiniband/hw/cxgb4/iw_cxgb4.h | 18 +++++++++++++ drivers/infiniband/hw/cxgb4/provider.c | 20 ++++++++++++--- drivers/infiniband/hw/cxgb4/qp.c | 35 +++++++++++++++++++------- 4 files changed, 69 insertions(+), 13 deletions(-) diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c index b99dc9e0ffb2..b85a1a983e07 100644 --- a/drivers/infiniband/hw/cxgb4/device.c +++ b/drivers/infiniband/hw/cxgb4/device.c @@ -848,9 +848,17 @@ static int c4iw_rdev_open(struct c4iw_rdev *rdev) } } + rdev->free_workq = create_singlethread_workqueue("iw_cxgb4_free"); + if (!rdev->free_workq) { + err = -ENOMEM; + goto err_free_status_page; + } + rdev->status_page->db_off = 0; return 0; +err_free_status_page: + free_page((unsigned long)rdev->status_page); destroy_ocqp_pool: c4iw_ocqp_pool_destroy(rdev); destroy_rqtpool: @@ -864,6 +872,7 @@ destroy_resource: static void c4iw_rdev_close(struct c4iw_rdev *rdev) { + destroy_workqueue(rdev->free_workq); kfree(rdev->wr_log); free_page((unsigned long)rdev->status_page); c4iw_pblpool_destroy(rdev); diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h index 4788e1a46fde..7d540667dad2 100644 --- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h +++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h @@ -45,6 +45,7 @@ #include #include #include +#include #include @@ -107,6 +108,7 @@ struct c4iw_dev_ucontext { struct list_head qpids; struct list_head cqids; struct mutex lock; + struct kref kref; }; enum c4iw_rdev_flags { @@ -183,6 +185,7 @@ struct c4iw_rdev { atomic_t wr_log_idx; struct wr_log_entry *wr_log; int wr_log_size; + struct workqueue_struct *free_workq; }; static inline int c4iw_fatal_error(struct c4iw_rdev *rdev) @@ -482,6 +485,8 @@ struct c4iw_qp { int sq_sig_all; struct completion rq_drained; struct completion sq_drained; + struct work_struct free_work; + struct c4iw_ucontext *ucontext; }; static inline struct c4iw_qp *to_c4iw_qp(struct ib_qp *ibqp) @@ -495,6 +500,7 @@ struct c4iw_ucontext { u32 key; spinlock_t mmap_lock; struct list_head mmaps; + struct kref kref; }; static inline struct c4iw_ucontext *to_c4iw_ucontext(struct ib_ucontext *c) @@ -502,6 +508,18 @@ static inline struct c4iw_ucontext *to_c4iw_ucontext(struct ib_ucontext *c) return container_of(c, struct c4iw_ucontext, ibucontext); } +void _c4iw_free_ucontext(struct kref *kref); + +static inline void c4iw_put_ucontext(struct c4iw_ucontext *ucontext) +{ + kref_put(&ucontext->kref, _c4iw_free_ucontext); +} + +static inline void c4iw_get_ucontext(struct c4iw_ucontext *ucontext) +{ + kref_get(&ucontext->kref); +} + struct c4iw_mm_entry { struct list_head entry; u64 addr; diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c index 645e606a17c5..8278ba06f995 100644 --- a/drivers/infiniband/hw/cxgb4/provider.c +++ b/drivers/infiniband/hw/cxgb4/provider.c @@ -91,17 +91,28 @@ static int c4iw_process_mad(struct ib_device *ibdev, int mad_flags, return -ENOSYS; } -static int c4iw_dealloc_ucontext(struct ib_ucontext *context) +void _c4iw_free_ucontext(struct kref *kref) { - struct c4iw_dev *rhp = to_c4iw_dev(context->device); - struct c4iw_ucontext *ucontext = to_c4iw_ucontext(context); + struct c4iw_ucontext *ucontext; + struct c4iw_dev *rhp; struct c4iw_mm_entry *mm, *tmp; - PDBG("%s context %p\n", __func__, context); + ucontext = container_of(kref, struct c4iw_ucontext, kref); + rhp = to_c4iw_dev(ucontext->ibucontext.device); + + PDBG("%s ucontext %p\n", __func__, ucontext); list_for_each_entry_safe(mm, tmp, &ucontext->mmaps, entry) kfree(mm); c4iw_release_dev_ucontext(&rhp->rdev, &ucontext->uctx); kfree(ucontext); +} + +static int c4iw_dealloc_ucontext(struct ib_ucontext *context) +{ + struct c4iw_ucontext *ucontext = to_c4iw_ucontext(context); + + PDBG("%s context %p\n", __func__, context); + c4iw_put_ucontext(ucontext); return 0; } @@ -125,6 +136,7 @@ static struct ib_ucontext *c4iw_alloc_ucontext(struct ib_device *ibdev, c4iw_init_dev_ucontext(&rhp->rdev, &context->uctx); INIT_LIST_HEAD(&context->mmaps); spin_lock_init(&context->mmap_lock); + kref_init(&context->kref); if (udata->outlen < sizeof(uresp) - sizeof(uresp.reserved)) { if (!warned++) diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c index b7ac97b27c88..bb0fde6e2047 100644 --- a/drivers/infiniband/hw/cxgb4/qp.c +++ b/drivers/infiniband/hw/cxgb4/qp.c @@ -714,13 +714,32 @@ static int build_inv_stag(union t4_wr *wqe, struct ib_send_wr *wr, u8 *len16) return 0; } -static void _free_qp(struct kref *kref) +static void free_qp_work(struct work_struct *work) +{ + struct c4iw_ucontext *ucontext; + struct c4iw_qp *qhp; + struct c4iw_dev *rhp; + + qhp = container_of(work, struct c4iw_qp, free_work); + ucontext = qhp->ucontext; + rhp = qhp->rhp; + + PDBG("%s qhp %p ucontext %p\n", __func__, qhp, ucontext); + destroy_qp(&rhp->rdev, &qhp->wq, + ucontext ? &ucontext->uctx : &rhp->rdev.uctx); + + if (ucontext) + c4iw_put_ucontext(ucontext); + kfree(qhp); +} + +static void queue_qp_free(struct kref *kref) { struct c4iw_qp *qhp; qhp = container_of(kref, struct c4iw_qp, kref); PDBG("%s qhp %p\n", __func__, qhp); - kfree(qhp); + queue_work(qhp->rhp->rdev.free_workq, &qhp->free_work); } void c4iw_qp_add_ref(struct ib_qp *qp) @@ -732,7 +751,7 @@ void c4iw_qp_add_ref(struct ib_qp *qp) void c4iw_qp_rem_ref(struct ib_qp *qp) { PDBG("%s ib_qp %p\n", __func__, qp); - kref_put(&to_c4iw_qp(qp)->kref, _free_qp); + kref_put(&to_c4iw_qp(qp)->kref, queue_qp_free); } static void add_to_fc_list(struct list_head *head, struct list_head *entry) @@ -1642,7 +1661,6 @@ int c4iw_destroy_qp(struct ib_qp *ib_qp) struct c4iw_dev *rhp; struct c4iw_qp *qhp; struct c4iw_qp_attributes attrs; - struct c4iw_ucontext *ucontext; qhp = to_c4iw_qp(ib_qp); rhp = qhp->rhp; @@ -1662,11 +1680,6 @@ int c4iw_destroy_qp(struct ib_qp *ib_qp) spin_unlock_irq(&rhp->lock); free_ird(rhp, qhp->attr.max_ird); - ucontext = ib_qp->uobject ? - to_c4iw_ucontext(ib_qp->uobject->context) : NULL; - destroy_qp(&rhp->rdev, &qhp->wq, - ucontext ? &ucontext->uctx : &rhp->rdev.uctx); - c4iw_qp_rem_ref(ib_qp); PDBG("%s ib_qp %p qpid 0x%0x\n", __func__, ib_qp, qhp->wq.sq.qid); @@ -1767,6 +1780,7 @@ struct ib_qp *c4iw_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *attrs, mutex_init(&qhp->mutex); init_waitqueue_head(&qhp->wait); kref_init(&qhp->kref); + INIT_WORK(&qhp->free_work, free_qp_work); ret = insert_handle(rhp, &rhp->qpidr, qhp, qhp->wq.sq.qid); if (ret) @@ -1853,6 +1867,9 @@ struct ib_qp *c4iw_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *attrs, ma_sync_key_mm->len = PAGE_SIZE; insert_mmap(ucontext, ma_sync_key_mm); } + + c4iw_get_ucontext(ucontext); + qhp->ucontext = ucontext; } qhp->ibqp.qp_num = qhp->wq.sq.qid; init_timer(&(qhp->timer)); From fe6531075e1dd8a7784bf0450186be1380eafb86 Mon Sep 17 00:00:00 2001 From: Max Kellermann Date: Thu, 15 Dec 2016 19:51:07 -0200 Subject: [PATCH 043/119] pctv452e: move buffer to heap, no mutex commit 48775cb73c2e26b7ca9d679875a6e570c8b8e124 upstream. commit 73d5c5c864f4 ("[media] pctv452e: don't do DMA on stack") caused a NULL pointer dereference which occurs when dvb_usb_init() calls dvb_usb_device_power_ctrl() for the first time, before the frontend has been attached. It also caused a recursive deadlock because tt3650_ci_msg_locked() has already locked the mutex. So, partially revert it, but move the buffer to the heap (DMA capable), not to the stack (may not be DMA capable). Instead of sharing one buffer which needs mutex protection, do a new heap allocation for each call. Fixes: commit 73d5c5c864f4 ("[media] pctv452e: don't do DMA on stack") Signed-off-by: Max Kellermann Signed-off-by: Mauro Carvalho Chehab Signed-off-by: Greg Kroah-Hartman --- drivers/media/usb/dvb-usb/pctv452e.c | 133 +++++++++++++++------------ 1 file changed, 72 insertions(+), 61 deletions(-) diff --git a/drivers/media/usb/dvb-usb/pctv452e.c b/drivers/media/usb/dvb-usb/pctv452e.c index 07fa08be9e99..d54ebe7e0215 100644 --- a/drivers/media/usb/dvb-usb/pctv452e.c +++ b/drivers/media/usb/dvb-usb/pctv452e.c @@ -97,14 +97,13 @@ struct pctv452e_state { u8 c; /* transaction counter, wraps around... */ u8 initialized; /* set to 1 if 0x15 has been sent */ u16 last_rc_key; - - unsigned char data[80]; }; static int tt3650_ci_msg(struct dvb_usb_device *d, u8 cmd, u8 *data, unsigned int write_len, unsigned int read_len) { struct pctv452e_state *state = (struct pctv452e_state *)d->priv; + u8 *buf; u8 id; unsigned int rlen; int ret; @@ -114,36 +113,39 @@ static int tt3650_ci_msg(struct dvb_usb_device *d, u8 cmd, u8 *data, return -EIO; } - mutex_lock(&state->ca_mutex); + buf = kmalloc(64, GFP_KERNEL); + if (!buf) + return -ENOMEM; + id = state->c++; - state->data[0] = SYNC_BYTE_OUT; - state->data[1] = id; - state->data[2] = cmd; - state->data[3] = write_len; + buf[0] = SYNC_BYTE_OUT; + buf[1] = id; + buf[2] = cmd; + buf[3] = write_len; - memcpy(state->data + 4, data, write_len); + memcpy(buf + 4, data, write_len); rlen = (read_len > 0) ? 64 : 0; - ret = dvb_usb_generic_rw(d, state->data, 4 + write_len, - state->data, rlen, /* delay_ms */ 0); + ret = dvb_usb_generic_rw(d, buf, 4 + write_len, + buf, rlen, /* delay_ms */ 0); if (0 != ret) goto failed; ret = -EIO; - if (SYNC_BYTE_IN != state->data[0] || id != state->data[1]) + if (SYNC_BYTE_IN != buf[0] || id != buf[1]) goto failed; - memcpy(data, state->data + 4, read_len); + memcpy(data, buf + 4, read_len); - mutex_unlock(&state->ca_mutex); + kfree(buf); return 0; failed: err("CI error %d; %02X %02X %02X -> %*ph.", - ret, SYNC_BYTE_OUT, id, cmd, 3, state->data); + ret, SYNC_BYTE_OUT, id, cmd, 3, buf); - mutex_unlock(&state->ca_mutex); + kfree(buf); return ret; } @@ -410,53 +412,57 @@ static int pctv452e_i2c_msg(struct dvb_usb_device *d, u8 addr, u8 *rcv_buf, u8 rcv_len) { struct pctv452e_state *state = (struct pctv452e_state *)d->priv; + u8 *buf; u8 id; int ret; - mutex_lock(&state->ca_mutex); + buf = kmalloc(64, GFP_KERNEL); + if (!buf) + return -ENOMEM; + id = state->c++; ret = -EINVAL; if (snd_len > 64 - 7 || rcv_len > 64 - 7) goto failed; - state->data[0] = SYNC_BYTE_OUT; - state->data[1] = id; - state->data[2] = PCTV_CMD_I2C; - state->data[3] = snd_len + 3; - state->data[4] = addr << 1; - state->data[5] = snd_len; - state->data[6] = rcv_len; + buf[0] = SYNC_BYTE_OUT; + buf[1] = id; + buf[2] = PCTV_CMD_I2C; + buf[3] = snd_len + 3; + buf[4] = addr << 1; + buf[5] = snd_len; + buf[6] = rcv_len; - memcpy(state->data + 7, snd_buf, snd_len); + memcpy(buf + 7, snd_buf, snd_len); - ret = dvb_usb_generic_rw(d, state->data, 7 + snd_len, - state->data, /* rcv_len */ 64, + ret = dvb_usb_generic_rw(d, buf, 7 + snd_len, + buf, /* rcv_len */ 64, /* delay_ms */ 0); if (ret < 0) goto failed; /* TT USB protocol error. */ ret = -EIO; - if (SYNC_BYTE_IN != state->data[0] || id != state->data[1]) + if (SYNC_BYTE_IN != buf[0] || id != buf[1]) goto failed; /* I2C device didn't respond as expected. */ ret = -EREMOTEIO; - if (state->data[5] < snd_len || state->data[6] < rcv_len) + if (buf[5] < snd_len || buf[6] < rcv_len) goto failed; - memcpy(rcv_buf, state->data + 7, rcv_len); - mutex_unlock(&state->ca_mutex); + memcpy(rcv_buf, buf + 7, rcv_len); + kfree(buf); return rcv_len; failed: err("I2C error %d; %02X %02X %02X %02X %02X -> %*ph", ret, SYNC_BYTE_OUT, id, addr << 1, snd_len, rcv_len, - 7, state->data); + 7, buf); - mutex_unlock(&state->ca_mutex); + kfree(buf); return ret; } @@ -505,7 +511,7 @@ static u32 pctv452e_i2c_func(struct i2c_adapter *adapter) static int pctv452e_power_ctrl(struct dvb_usb_device *d, int i) { struct pctv452e_state *state = (struct pctv452e_state *)d->priv; - u8 *rx; + u8 *b0, *rx; int ret; info("%s: %d\n", __func__, i); @@ -516,11 +522,12 @@ static int pctv452e_power_ctrl(struct dvb_usb_device *d, int i) if (state->initialized) return 0; - rx = kmalloc(PCTV_ANSWER_LEN, GFP_KERNEL); - if (!rx) + b0 = kmalloc(5 + PCTV_ANSWER_LEN, GFP_KERNEL); + if (!b0) return -ENOMEM; - mutex_lock(&state->ca_mutex); + rx = b0 + 5; + /* hmm where shoud this should go? */ ret = usb_set_interface(d->udev, 0, ISOC_INTERFACE_ALTERNATIVE); if (ret != 0) @@ -528,66 +535,70 @@ static int pctv452e_power_ctrl(struct dvb_usb_device *d, int i) __func__, ret); /* this is a one-time initialization, dont know where to put */ - state->data[0] = 0xaa; - state->data[1] = state->c++; - state->data[2] = PCTV_CMD_RESET; - state->data[3] = 1; - state->data[4] = 0; + b0[0] = 0xaa; + b0[1] = state->c++; + b0[2] = PCTV_CMD_RESET; + b0[3] = 1; + b0[4] = 0; /* reset board */ - ret = dvb_usb_generic_rw(d, state->data, 5, rx, PCTV_ANSWER_LEN, 0); + ret = dvb_usb_generic_rw(d, b0, 5, rx, PCTV_ANSWER_LEN, 0); if (ret) goto ret; - state->data[1] = state->c++; - state->data[4] = 1; + b0[1] = state->c++; + b0[4] = 1; /* reset board (again?) */ - ret = dvb_usb_generic_rw(d, state->data, 5, rx, PCTV_ANSWER_LEN, 0); + ret = dvb_usb_generic_rw(d, b0, 5, rx, PCTV_ANSWER_LEN, 0); if (ret) goto ret; state->initialized = 1; ret: - mutex_unlock(&state->ca_mutex); - kfree(rx); + kfree(b0); return ret; } static int pctv452e_rc_query(struct dvb_usb_device *d) { struct pctv452e_state *state = (struct pctv452e_state *)d->priv; + u8 *b, *rx; int ret, i; u8 id; - mutex_lock(&state->ca_mutex); + b = kmalloc(CMD_BUFFER_SIZE + PCTV_ANSWER_LEN, GFP_KERNEL); + if (!b) + return -ENOMEM; + + rx = b + CMD_BUFFER_SIZE; + id = state->c++; /* prepare command header */ - state->data[0] = SYNC_BYTE_OUT; - state->data[1] = id; - state->data[2] = PCTV_CMD_IR; - state->data[3] = 0; + b[0] = SYNC_BYTE_OUT; + b[1] = id; + b[2] = PCTV_CMD_IR; + b[3] = 0; /* send ir request */ - ret = dvb_usb_generic_rw(d, state->data, 4, - state->data, PCTV_ANSWER_LEN, 0); + ret = dvb_usb_generic_rw(d, b, 4, rx, PCTV_ANSWER_LEN, 0); if (ret != 0) goto ret; if (debug > 3) { - info("%s: read: %2d: %*ph: ", __func__, ret, 3, state->data); - for (i = 0; (i < state->data[3]) && ((i + 3) < PCTV_ANSWER_LEN); i++) - info(" %02x", state->data[i + 3]); + info("%s: read: %2d: %*ph: ", __func__, ret, 3, rx); + for (i = 0; (i < rx[3]) && ((i+3) < PCTV_ANSWER_LEN); i++) + info(" %02x", rx[i+3]); info("\n"); } - if ((state->data[3] == 9) && (state->data[12] & 0x01)) { + if ((rx[3] == 9) && (rx[12] & 0x01)) { /* got a "press" event */ - state->last_rc_key = RC_SCANCODE_RC5(state->data[7], state->data[6]); + state->last_rc_key = RC_SCANCODE_RC5(rx[7], rx[6]); if (debug > 2) info("%s: cmd=0x%02x sys=0x%02x\n", - __func__, state->data[6], state->data[7]); + __func__, rx[6], rx[7]); rc_keydown(d->rc_dev, RC_TYPE_RC5, state->last_rc_key, 0); } else if (state->last_rc_key) { @@ -595,7 +606,7 @@ static int pctv452e_rc_query(struct dvb_usb_device *d) state->last_rc_key = 0; } ret: - mutex_unlock(&state->ca_mutex); + kfree(b); return ret; } From d25f9bfeb62f812e8aa7eded53df8155bff25ee5 Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Fri, 9 Dec 2016 09:47:17 -0200 Subject: [PATCH 044/119] v4l: tvp5150: Reset device at probe time, not in get/set format handlers commit aff808e813fc2d311137754165cf53d4ee6ddcc2 upstream. The tvp5150 doesn't support format setting through the subdev pad API and thus implements the set format handler as a get format operation. The single handler, tvp5150_fill_fmt(), resets the device by calling tvp5150_reset(). This causes malfunction as the device can be reset at will, possibly from userspace when the subdev userspace API is enabled. The reset call was added in commit ec2c4f3f93cb ("[media] media: tvp5150: Add mbus_fmt callbacks"), probably as an attempt to set the device to a known state before detecting the current TV standard. However, the get format handler doesn't access the hardware to get the TV standard since commit 963ddc63e20d ("[media] media: tvp5150: Add cropping support"). There is thus no need to reset the device when getting the format. However, removing the tvp5150_reset() from the get/set format handlers results in the function not being called at all if the bridge driver doesn't use the .reset() operation. The operation is nowadays abused and shouldn't be used, so shouldn't expect bridge drivers to call it. To make sure the device is properly initialize, move the reset call from the format handlers to the probe function. Signed-off-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab Signed-off-by: Greg Kroah-Hartman --- drivers/media/i2c/tvp5150.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/media/i2c/tvp5150.c b/drivers/media/i2c/tvp5150.c index 7268e706e216..7fa359ea3031 100644 --- a/drivers/media/i2c/tvp5150.c +++ b/drivers/media/i2c/tvp5150.c @@ -858,8 +858,6 @@ static int tvp5150_fill_fmt(struct v4l2_subdev *sd, f = &format->format; - tvp5150_reset(sd, 0); - f->width = decoder->rect.width; f->height = decoder->rect.height / 2; @@ -1521,7 +1519,6 @@ static int tvp5150_probe(struct i2c_client *c, res = core->hdl.error; goto err; } - v4l2_ctrl_handler_setup(&core->hdl); /* Default is no cropping */ core->rect.top = 0; @@ -1532,6 +1529,8 @@ static int tvp5150_probe(struct i2c_client *c, core->rect.left = 0; core->rect.width = TVP5150_H_MAX; + tvp5150_reset(sd, 0); /* Calls v4l2_ctrl_handler_setup() */ + res = v4l2_async_register_subdev(sd); if (res < 0) goto err; From 11e5015ae1d12473f12baa1e8cc66849be019c3a Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Fri, 9 Dec 2016 09:47:18 -0200 Subject: [PATCH 045/119] v4l: tvp5150: Fix comment regarding output pin muxing commit b4b2de386bbb6589d81596999d4a924928dc119b upstream. The FID/GLCO/VLK/HVLK and INTREQ/GPCL/VBLK pins are muxed differently depending on whether the input is an S-Video or composite signal. The comment that explains the logic doesn't reflect the code. It appears that the comment is incorrect, as disabling the output data bus in composite mode makes no sense. Update the comment to match the code. While at it define macros for the MISC_CTL register bits, the code is too confusing with numerical values. Signed-off-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab Signed-off-by: Greg Kroah-Hartman --- drivers/media/i2c/tvp5150.c | 24 +++++++++++++++++------- drivers/media/i2c/tvp5150_reg.h | 9 +++++++++ 2 files changed, 26 insertions(+), 7 deletions(-) diff --git a/drivers/media/i2c/tvp5150.c b/drivers/media/i2c/tvp5150.c index 7fa359ea3031..e88d2d25b37e 100644 --- a/drivers/media/i2c/tvp5150.c +++ b/drivers/media/i2c/tvp5150.c @@ -288,8 +288,12 @@ static inline void tvp5150_selmux(struct v4l2_subdev *sd) tvp5150_write(sd, TVP5150_OP_MODE_CTL, opmode); tvp5150_write(sd, TVP5150_VD_IN_SRC_SEL_1, input); - /* Svideo should enable YCrCb output and disable GPCL output - * For Composite and TV, it should be the reverse + /* + * Setup the FID/GLCO/VLK/HVLK and INTREQ/GPCL/VBLK output signals. For + * S-Video we output the vertical lock (VLK) signal on FID/GLCO/VLK/HVLK + * and set INTREQ/GPCL/VBLK to logic 0. For composite we output the + * field indicator (FID) signal on FID/GLCO/VLK/HVLK and set + * INTREQ/GPCL/VBLK to logic 1. */ val = tvp5150_read(sd, TVP5150_MISC_CTL); if (val < 0) { @@ -298,9 +302,9 @@ static inline void tvp5150_selmux(struct v4l2_subdev *sd) } if (decoder->input == TVP5150_SVIDEO) - val = (val & ~0x40) | 0x10; + val = (val & ~TVP5150_MISC_CTL_GPCL) | TVP5150_MISC_CTL_HVLK; else - val = (val & ~0x10) | 0x40; + val = (val & ~TVP5150_MISC_CTL_HVLK) | TVP5150_MISC_CTL_GPCL; tvp5150_write(sd, TVP5150_MISC_CTL, val); }; @@ -452,7 +456,12 @@ static const struct i2c_reg_value tvp5150_init_enable[] = { },{ /* Automatic offset and AGC enabled */ TVP5150_ANAL_CHL_CTL, 0x15 },{ /* Activate YCrCb output 0x9 or 0xd ? */ - TVP5150_MISC_CTL, 0x6f + TVP5150_MISC_CTL, TVP5150_MISC_CTL_GPCL | + TVP5150_MISC_CTL_INTREQ_OE | + TVP5150_MISC_CTL_YCBCR_OE | + TVP5150_MISC_CTL_SYNC_OE | + TVP5150_MISC_CTL_VBLANK | + TVP5150_MISC_CTL_CLOCK_OE, },{ /* Activates video std autodetection for all standards */ TVP5150_AUTOSW_MSK, 0x0 },{ /* Default format: 0x47. For 4:2:2: 0x40 */ @@ -1047,11 +1056,12 @@ static int tvp5150_s_stream(struct v4l2_subdev *sd, int enable) { struct tvp5150 *decoder = to_tvp5150(sd); /* Output format: 8-bit ITU-R BT.656 with embedded syncs */ - int val = 0x09; + int val = TVP5150_MISC_CTL_YCBCR_OE | TVP5150_MISC_CTL_CLOCK_OE; /* Output format: 8-bit 4:2:2 YUV with discrete sync */ if (decoder->mbus_type == V4L2_MBUS_PARALLEL) - val = 0x0d; + val = TVP5150_MISC_CTL_YCBCR_OE | TVP5150_MISC_CTL_SYNC_OE + | TVP5150_MISC_CTL_CLOCK_OE; /* Initializes TVP5150 to its default values */ /* # set PCLK (27MHz) */ diff --git a/drivers/media/i2c/tvp5150_reg.h b/drivers/media/i2c/tvp5150_reg.h index 25a994944918..30a48c28d05a 100644 --- a/drivers/media/i2c/tvp5150_reg.h +++ b/drivers/media/i2c/tvp5150_reg.h @@ -9,6 +9,15 @@ #define TVP5150_ANAL_CHL_CTL 0x01 /* Analog channel controls */ #define TVP5150_OP_MODE_CTL 0x02 /* Operation mode controls */ #define TVP5150_MISC_CTL 0x03 /* Miscellaneous controls */ +#define TVP5150_MISC_CTL_VBLK_GPCL BIT(7) +#define TVP5150_MISC_CTL_GPCL BIT(6) +#define TVP5150_MISC_CTL_INTREQ_OE BIT(5) +#define TVP5150_MISC_CTL_HVLK BIT(4) +#define TVP5150_MISC_CTL_YCBCR_OE BIT(3) +#define TVP5150_MISC_CTL_SYNC_OE BIT(2) +#define TVP5150_MISC_CTL_VBLANK BIT(1) +#define TVP5150_MISC_CTL_CLOCK_OE BIT(0) + #define TVP5150_AUTOSW_MSK 0x04 /* Autoswitch mask: TVP5150A / TVP5150AM */ /* Reserved 05h */ From 1b8ff91ae545d36337de459c7195f6f369dc776d Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Fri, 9 Dec 2016 09:47:19 -0200 Subject: [PATCH 046/119] v4l: tvp5150: Don't override output pinmuxing at stream on/off time commit 79d6205a3f741c9fb89cfc47dfa0eddb1526726d upstream. The s_stream() handler incorrectly writes the whole MISC_CTL register to enable or disable the outputs, overriding the output pinmuxing configuration. Fix it to only touch the output enable bits. The CONF_SHARED_PIN register is also written by the same function, resulting in muxing the INTREQ signal instead of the VBLK/GPCL signal on the INTREQ/GPCL/VBLK pin. As the driver doesn't support interrupts this is obviously incorrect, and breaks operation on other devices. Fix it by removing the write. Signed-off-by: Laurent Pinchart Signed-off-by: Mauro Carvalho Chehab Signed-off-by: Greg Kroah-Hartman --- drivers/media/i2c/tvp5150.c | 31 ++++++++++++++++++------------- 1 file changed, 18 insertions(+), 13 deletions(-) diff --git a/drivers/media/i2c/tvp5150.c b/drivers/media/i2c/tvp5150.c index e88d2d25b37e..59aa4dafb60b 100644 --- a/drivers/media/i2c/tvp5150.c +++ b/drivers/media/i2c/tvp5150.c @@ -1055,22 +1055,27 @@ static const struct media_entity_operations tvp5150_sd_media_ops = { static int tvp5150_s_stream(struct v4l2_subdev *sd, int enable) { struct tvp5150 *decoder = to_tvp5150(sd); - /* Output format: 8-bit ITU-R BT.656 with embedded syncs */ - int val = TVP5150_MISC_CTL_YCBCR_OE | TVP5150_MISC_CTL_CLOCK_OE; + int val; - /* Output format: 8-bit 4:2:2 YUV with discrete sync */ - if (decoder->mbus_type == V4L2_MBUS_PARALLEL) - val = TVP5150_MISC_CTL_YCBCR_OE | TVP5150_MISC_CTL_SYNC_OE - | TVP5150_MISC_CTL_CLOCK_OE; + /* Enable or disable the video output signals. */ + val = tvp5150_read(sd, TVP5150_MISC_CTL); + if (val < 0) + return val; - /* Initializes TVP5150 to its default values */ - /* # set PCLK (27MHz) */ - tvp5150_write(sd, TVP5150_CONF_SHARED_PIN, 0x00); + val &= ~(TVP5150_MISC_CTL_YCBCR_OE | TVP5150_MISC_CTL_SYNC_OE | + TVP5150_MISC_CTL_CLOCK_OE); - if (enable) - tvp5150_write(sd, TVP5150_MISC_CTL, val); - else - tvp5150_write(sd, TVP5150_MISC_CTL, 0x00); + if (enable) { + /* + * Enable the YCbCr and clock outputs. In discrete sync mode + * (non-BT.656) additionally enable the the sync outputs. + */ + val |= TVP5150_MISC_CTL_YCBCR_OE | TVP5150_MISC_CTL_CLOCK_OE; + if (decoder->mbus_type == V4L2_MBUS_PARALLEL) + val |= TVP5150_MISC_CTL_SYNC_OE; + } + + tvp5150_write(sd, TVP5150_MISC_CTL, val); return 0; } From 9913aca2d5879a474f4ecb07c081ef1d0f7b3cf0 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Thu, 5 Jan 2017 15:59:40 +0000 Subject: [PATCH 047/119] drm/i915: Clear ret before unbinding in i915_gem_evict_something() commit e88893fea17996018b2d68a22e677ea04f3baadf upstream. Missed when rebasing patches, I failed to set ret to zero before starting the unbind loop (which depends upon ret being zero). Reported-by: Matthew Auld Fixes: 9332f3b1b99a ("drm/i915: Combine loops within i915_gem_evict_something") Signed-off-by: Chris Wilson Cc: Matthew Auld Link: http://patchwork.freedesktop.org/patch/msgid/20170105155940.10033-1-chris@chris-wilson.co.uk Reviewed-by: Matthew Auld (cherry picked from commit 121dfbb2a2ef1c5f49e15c38ccc47ff0beb59446) Signed-off-by: Jani Nikula Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/i915_gem_evict.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index 5b6f81c1dbca..7467355e4a18 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -194,6 +194,7 @@ found: } /* Unbinding will emit any required flushes */ + ret = 0; while (!list_empty(&eviction_list)) { vma = list_first_entry(&eviction_list, struct i915_vma, From 2cff678d0bb506595f7ca765c6e3b3ff0130f7a9 Mon Sep 17 00:00:00 2001 From: Clint Taylor Date: Wed, 18 Jan 2017 13:38:43 -0800 Subject: [PATCH 048/119] drm/i915: prevent crash with .disable_display parameter commit 27892bbdc9233f33bf0f44e08aab8f12e0dec142 upstream. The .disable_display parameter was causing a fatal crash when fbdev was dereferenced during driver init. V1: protection in i915_drv.c V2: Moved protection to intel_fbdev.c Fixes: 43cee314345a ("drm/i915/fbdev: Limit the global async-domain synchronization") Testcase: igt/drv_module_reload/basic-no-display Cc: Chris Wilson Signed-off-by: Clint Taylor Link: http://patchwork.freedesktop.org/patch/msgid/1484775523-29428-1-git-send-email-clinton.a.taylor@intel.com Reviewed-by: Chris Wilson Cc: Lukas Wunner Cc: Daniel Vetter Cc: Jani Nikula Signed-off-by: Chris Wilson (cherry picked from commit 5b8cd0755f8a06a851c436a013e7be0823fb155a) Signed-off-by: Jani Nikula Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/intel_fbdev.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c index b7098f98bb67..9127e57f383c 100644 --- a/drivers/gpu/drm/i915/intel_fbdev.c +++ b/drivers/gpu/drm/i915/intel_fbdev.c @@ -745,6 +745,9 @@ void intel_fbdev_initial_config_async(struct drm_device *dev) { struct intel_fbdev *ifbdev = to_i915(dev)->fbdev; + if (!ifbdev) + return; + ifbdev->cookie = async_schedule(intel_fbdev_initial_config, ifbdev); } From 8d7c76ae613ed9efd4548e85b894794a562dada4 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Jan 2017 16:28:42 +0200 Subject: [PATCH 049/119] drm/i915: Don't leak edid in intel_crt_detect_ddc() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit c34f078675f505c4437919bb1897b1351f16a050 upstream. In the path where intel_crt_detect_ddc() detects a CRT, if would return true without freeing the edid. Fixes: a2bd1f541f19 ("drm/i915: check whether we actually received an edid in detect_ddc") Cc: Chris Wilson Cc: Daniel Vetter Cc: Daniel Vetter Cc: Jani Nikula Cc: intel-gfx@lists.freedesktop.org Signed-off-by: Ander Conselvan de Oliveira Reviewed-by: Ville Syrjälä Reviewed-by: Jani Nikula Link: http://patchwork.freedesktop.org/patch/msgid/1484922525-6131-1-git-send-email-ander.conselvan.de.oliveira@intel.com (cherry picked from commit c96b63a6a7ac4bd670ec2e663793a9a31418b790) Signed-off-by: Jani Nikula Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/intel_crt.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index dfbcf16b41df..4149a0fbe8bd 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -499,6 +499,7 @@ static bool intel_crt_detect_ddc(struct drm_connector *connector) struct drm_i915_private *dev_priv = to_i915(crt->base.base.dev); struct edid *edid; struct i2c_adapter *i2c; + bool ret = false; BUG_ON(crt->base.type != INTEL_OUTPUT_ANALOG); @@ -515,17 +516,17 @@ static bool intel_crt_detect_ddc(struct drm_connector *connector) */ if (!is_digital) { DRM_DEBUG_KMS("CRT detected via DDC:0x50 [EDID]\n"); - return true; + ret = true; + } else { + DRM_DEBUG_KMS("CRT not detected via DDC:0x50 [EDID reports a digital panel]\n"); } - - DRM_DEBUG_KMS("CRT not detected via DDC:0x50 [EDID reports a digital panel]\n"); } else { DRM_DEBUG_KMS("CRT not detected via DDC:0x50 [no valid EDID found]\n"); } kfree(edid); - return false; + return ret; } static enum drm_connector_status From ff4956555513d0ad0088866c748a30bf10a61019 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Jan 2017 16:28:43 +0200 Subject: [PATCH 050/119] drm/i915: Don't init hpd polling for vlv and chv from runtime_suspend() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit 21d6e0bde50713922a6520ef84e5fd245b05d468 upstream. An error in the condition for avoiding the call to intel_hpd_poll_init() for valleyview and cherryview from intel_runtime_suspend() caused it to be called unconditionally. Fix it. Fixes: 19625e85c6ec ("drm/i915: Enable polling when we don't have hpd") Cc: Ville Syrjälä Cc: Daniel Vetter Cc: Lyude Cc: Daniel Vetter Cc: Jani Nikula Cc: intel-gfx@lists.freedesktop.org Signed-off-by: Ander Conselvan de Oliveira Reviewed-by: Ville Syrjälä Link: http://patchwork.freedesktop.org/patch/msgid/1484922525-6131-2-git-send-email-ander.conselvan.de.oliveira@intel.com (cherry picked from commit 04313b00b79405f86d815100f85c47a2ee5b8ca0) Signed-off-by: Jani Nikula Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/i915_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 18dfdd5c1b3b..670beebc32f6 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -2372,7 +2372,7 @@ static int intel_runtime_suspend(struct device *kdev) assert_forcewakes_inactive(dev_priv); - if (!IS_VALLEYVIEW(dev_priv) || !IS_CHERRYVIEW(dev_priv)) + if (!IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv)) intel_hpd_poll_init(dev_priv); DRM_DEBUG_KMS("Device suspended\n"); From 181478cdb75eb4d9865064b15cb93418c72759f2 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Jan 2017 16:28:44 +0200 Subject: [PATCH 051/119] drm/i915: Fix calculation of rotated x and y offsets for planar formats MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit 3781bd6e7d64d5f5bea9fdee11ab9460a700c0e4 upstream. Parameters tile_size, tile_width and tile_height were passed in the wrong order to _intel_adjust_tile_offset() when calculating the rotated offsets. This doesn't fix any user visible bug, since for packed formats new and old offset are the same and the rotated offsets are within a tile before they are fed to _intel_adjust_tile_offset(). In that case, the offsets are unchanged. That is not true for planar formats, but those are currently not supported. Fixes: 66a2d927cb0e ("drm/i915: Make intel_adjust_tile_offset() work for linear buffers") Cc: Ville Syrjälä Cc: Sivakumar Thulasimani Cc: Daniel Vetter Cc: Jani Nikula Cc: intel-gfx@lists.freedesktop.org Signed-off-by: Ander Conselvan de Oliveira Reviewed-by: Ville Syrjälä Link: http://patchwork.freedesktop.org/patch/msgid/1484922525-6131-3-git-send-email-ander.conselvan.de.oliveira@intel.com (cherry picked from commit 46a1bd289507dfcc428fb9daf65421ed6be6af8b) Signed-off-by: Jani Nikula Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/intel_display.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 3c9cb35e46ab..944e067a9f28 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2587,8 +2587,9 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv, * We only keep the x/y offsets, so push all of the * gtt offset into the x/y offsets. */ - _intel_adjust_tile_offset(&x, &y, tile_size, - tile_width, tile_height, pitch_tiles, + _intel_adjust_tile_offset(&x, &y, + tile_width, tile_height, + tile_size, pitch_tiles, gtt_offset_rotated * tile_size, 0); gtt_offset_rotated += rot_info->plane[i].width * rot_info->plane[i].height; From d7c3d7e453b70e9e9e81d8409e5f8c004597ec05 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Jan 2017 16:28:45 +0200 Subject: [PATCH 052/119] drm/i915: Check for NULL atomic state in intel_crtc_disable_noatomic() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit commit 6d1d427a4e24c403b4adf928d61994bdaa0ca03a upstream. In intel_crtc_disable_noatomic(), bail on a failure to allocate an atomic state to avoid a NULL pointer dereference. Fixes: 4a80655827af ("drm/i915: Pass atomic state to crtc enable/disable functions") Cc: Maarten Lankhorst Cc: Daniel Vetter Cc: Daniel Vetter Cc: Jani Nikula Cc: intel-gfx@lists.freedesktop.org Signed-off-by: Ander Conselvan de Oliveira Reviewed-by: Ville Syrjälä Link: http://patchwork.freedesktop.org/patch/msgid/1484922525-6131-4-git-send-email-ander.conselvan.de.oliveira@intel.com (cherry picked from commit 31bb2ef97ea9db343348f9b5ccaa9bb6f48fc655) Signed-off-by: Jani Nikula Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/intel_display.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 944e067a9f28..8079e5b380cb 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -6869,6 +6869,12 @@ static void intel_crtc_disable_noatomic(struct drm_crtc *crtc) } state = drm_atomic_state_alloc(crtc->dev); + if (!state) { + DRM_DEBUG_KMS("failed to disable [CRTC:%d:%s], out of memory", + crtc->base.id, crtc->name); + return; + } + state->acquire_ctx = crtc->dev->mode_config.acquire_ctx; /* Everything's already locked, -EDEADLK can't happen. */ From c81ee4ed9c0d49514254487eb037548dd5e25ef8 Mon Sep 17 00:00:00 2001 From: Kenneth Lee Date: Thu, 5 Jan 2017 15:00:05 +0800 Subject: [PATCH 053/119] IB/umem: Release pid in error and ODP flow commit 828f6fa65ce7e80f77f5ab12942e44eb3d9d174e upstream. 1. Release pid before enter odp flow 2. Release pid when fail to allocate memory Fixes: 87773dd56d54 ("IB: ib_umem_release() should decrement mm->pinned_vm from ib_umem_get") Fixes: 8ada2c1c0c1d ("IB/core: Add support for on demand paging regions") Signed-off-by: Kenneth Lee Reviewed-by: Haggai Eran Reviewed-by: Leon Romanovsky Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/core/umem.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/infiniband/core/umem.c b/drivers/infiniband/core/umem.c index 84b4eff90395..c22fde6207d1 100644 --- a/drivers/infiniband/core/umem.c +++ b/drivers/infiniband/core/umem.c @@ -134,6 +134,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, IB_ACCESS_REMOTE_ATOMIC | IB_ACCESS_MW_BIND)); if (access & IB_ACCESS_ON_DEMAND) { + put_pid(umem->pid); ret = ib_umem_odp_get(context, umem); if (ret) { kfree(umem); @@ -149,6 +150,7 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, page_list = (struct page **) __get_free_page(GFP_KERNEL); if (!page_list) { + put_pid(umem->pid); kfree(umem); return ERR_PTR(-ENOMEM); } From fed1e89a16e3684e1bd218e1d6a78f83b5ceb865 Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Thu, 19 Jan 2017 15:25:58 +0200 Subject: [PATCH 054/119] IB/rxe: Fix rxe dev insertion to rxe_dev_list commit f39f775218a7520e3700de2003c84a042c3b5972 upstream. The first argument of list_add_tail is the new item and the second is the head of the list. Fix the code to pass arguments in the right order, otherwise not all the rxe devices will be removed during teardown. Fixes: 8700e3e7c4857 ('Soft RoCE driver') Signed-off-by: Maor Gottlieb Reviewed-by: Moni Shoua Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/sw/rxe/rxe_net.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c index ffff5a54cb34..f4f3942ebbd1 100644 --- a/drivers/infiniband/sw/rxe/rxe_net.c +++ b/drivers/infiniband/sw/rxe/rxe_net.c @@ -554,7 +554,7 @@ struct rxe_dev *rxe_net_add(struct net_device *ndev) } spin_lock_bh(&dev_list_lock); - list_add_tail(&rxe_dev_list, &rxe->list); + list_add_tail(&rxe->list, &rxe_dev_list); spin_unlock_bh(&dev_list_lock); return rxe; } From 9a335996bfcfce1fd7fe7549fb1a679489774b99 Mon Sep 17 00:00:00 2001 From: Yonatan Cohen Date: Thu, 19 Jan 2017 15:25:59 +0200 Subject: [PATCH 055/119] IB/rxe: Prevent from completer to operate on non valid QP commit 2d4b21e0a2913612274a69a3ba1bfee4cffc6e77 upstream. On UD QP completer tasklet is scheduled for each packet sent. If it is followed by a destroy_qp(), the kernel panic will happen as the completer tries to operate on a destroyed QP. Fixes: 8700e3e7c485 ("Soft RoCE driver") Signed-off-by: Yonatan Cohen Reviewed-by: Moni Shoua Signed-off-by: Leon Romanovsky Signed-off-by: Doug Ledford Signed-off-by: Greg Kroah-Hartman --- drivers/infiniband/sw/rxe/rxe_qp.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c index 486d576e55bc..44b2108253bd 100644 --- a/drivers/infiniband/sw/rxe/rxe_qp.c +++ b/drivers/infiniband/sw/rxe/rxe_qp.c @@ -813,8 +813,7 @@ void rxe_qp_destroy(struct rxe_qp *qp) del_timer_sync(&qp->rnr_nak_timer); rxe_cleanup_task(&qp->req.task); - if (qp_type(qp) == IB_QPT_RC) - rxe_cleanup_task(&qp->comp.task); + rxe_cleanup_task(&qp->comp.task); /* flush out any receive wr's or pending requests */ __rxe_do_task(&qp->req.task); From 668a827a6c9daa88efb4464144a6bd7b2146c66c Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 9 Dec 2016 09:41:29 -0200 Subject: [PATCH 056/119] s5k4ecgx: select CRC32 helper commit c739c0a7c3c2472d7562b8f802cdce44d2597c8b upstream. A rare randconfig build failure shows up in this driver when the CRC32 helper is not there: drivers/media/built-in.o: In function `s5k4ecgx_s_power': s5k4ecgx.c:(.text+0x9eb4): undefined reference to `crc32_le' This adds the 'select' that all other users of this function have. Fixes: 8b99312b7214 ("[media] Add v4l2 subdev driver for S5K4ECGX sensor") Signed-off-by: Arnd Bergmann Signed-off-by: Mauro Carvalho Chehab Signed-off-by: Greg Kroah-Hartman --- drivers/media/i2c/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/media/i2c/Kconfig b/drivers/media/i2c/Kconfig index 2669b4bad910..5a27bffa02fb 100644 --- a/drivers/media/i2c/Kconfig +++ b/drivers/media/i2c/Kconfig @@ -655,6 +655,7 @@ config VIDEO_S5K6A3 config VIDEO_S5K4ECGX tristate "Samsung S5K4ECGX sensor support" depends on I2C && VIDEO_V4L2 && VIDEO_V4L2_SUBDEV_API + select CRC32 ---help--- This is a V4L2 sensor-level driver for Samsung S5K4ECGX 5M camera sensor with an embedded SoC image signal processor. From 35948ae694c0d6aceda6d41ca1b624206f702da1 Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Tue, 10 Jan 2017 17:31:56 +0300 Subject: [PATCH 057/119] pinctrl: broxton: Use correct PADCFGLOCK offset commit ecc8995363ee6231b32dad61c955b371b79cc4cf upstream. PADCFGLOCK (and PADCFGLOCK_TX) offset in Broxton actually starts at 0x060 and not 0x090 as used in the driver. Fix it to use the correct offset. Signed-off-by: Mika Westerberg Reviewed-by: Andy Shevchenko Signed-off-by: Linus Walleij Signed-off-by: Greg Kroah-Hartman --- drivers/pinctrl/intel/pinctrl-broxton.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pinctrl/intel/pinctrl-broxton.c b/drivers/pinctrl/intel/pinctrl-broxton.c index 59cb7a6fc5be..901b356b09d7 100644 --- a/drivers/pinctrl/intel/pinctrl-broxton.c +++ b/drivers/pinctrl/intel/pinctrl-broxton.c @@ -19,7 +19,7 @@ #define BXT_PAD_OWN 0x020 #define BXT_HOSTSW_OWN 0x080 -#define BXT_PADCFGLOCK 0x090 +#define BXT_PADCFGLOCK 0x060 #define BXT_GPI_IE 0x110 #define BXT_COMMUNITY(s, e) \ From 583eded5860b84344d91462796cc5e5d40bbc27b Mon Sep 17 00:00:00 2001 From: Masahiro Yamada Date: Tue, 17 Jan 2017 19:52:54 +0900 Subject: [PATCH 058/119] pinctrl: uniphier: fix Ethernet (RMII) pin-mux setting for LD20 commit df1539c25cce98e2ac69881958850c6535240707 upstream. Fix the pin-mux values for the MDC, MDIO, MDIO_INTL, PHYRSTL pins. Fixes: 1e359ab1285e ("pinctrl: uniphier: add Ethernet pin-mux settings") Signed-off-by: Masahiro Yamada Signed-off-by: Linus Walleij Signed-off-by: Greg Kroah-Hartman --- drivers/pinctrl/uniphier/pinctrl-uniphier-ld20.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pinctrl/uniphier/pinctrl-uniphier-ld20.c b/drivers/pinctrl/uniphier/pinctrl-uniphier-ld20.c index aa8bd9794683..96686336e3a3 100644 --- a/drivers/pinctrl/uniphier/pinctrl-uniphier-ld20.c +++ b/drivers/pinctrl/uniphier/pinctrl-uniphier-ld20.c @@ -561,7 +561,7 @@ static const int ether_rgmii_muxvals[] = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}; static const unsigned ether_rmii_pins[] = {30, 31, 32, 33, 34, 35, 36, 37, 39, 41, 42, 45}; -static const int ether_rmii_muxvals[] = {1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}; +static const int ether_rmii_muxvals[] = {0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1}; static const unsigned i2c0_pins[] = {63, 64}; static const int i2c0_muxvals[] = {0, 0}; static const unsigned i2c1_pins[] = {65, 66}; From da1fdb8456ac79533aa09ce61b99a84ac7ca57b9 Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Tue, 10 Jan 2017 16:38:52 +0200 Subject: [PATCH 059/119] pinctrl: baytrail: Rectify debounce support commit 04ff5a095d662e0879f0eb04b9247e092210aeff upstream. The commit 658b476c742f ("pinctrl: baytrail: Add debounce configuration") implements debounce for Baytrail pin control, but seems wasn't tested properly. The register which keeps debounce value is separated from the configuration one. Writing wrong values to the latter will guarantee wrong behaviour of the driver and even might break something physically. Besides above there is missed case how to disable it, which is actually done through the bit in configuration register. Rectify implementation here by using proper register for debounce value. Fixes: 658b476c742f ("pinctrl: baytrail: Add debounce configuration") Cc: Cristina Ciocan Signed-off-by: Andy Shevchenko Acked-by: Mika Westerberg Signed-off-by: Linus Walleij Signed-off-by: Greg Kroah-Hartman --- drivers/pinctrl/intel/pinctrl-baytrail.c | 28 ++++++++++++++---------- 1 file changed, 17 insertions(+), 11 deletions(-) diff --git a/drivers/pinctrl/intel/pinctrl-baytrail.c b/drivers/pinctrl/intel/pinctrl-baytrail.c index 71bbeb9321ba..079015385fd8 100644 --- a/drivers/pinctrl/intel/pinctrl-baytrail.c +++ b/drivers/pinctrl/intel/pinctrl-baytrail.c @@ -1092,6 +1092,7 @@ static int byt_pin_config_get(struct pinctrl_dev *pctl_dev, unsigned int offset, enum pin_config_param param = pinconf_to_config_param(*config); void __iomem *conf_reg = byt_gpio_reg(vg, offset, BYT_CONF0_REG); void __iomem *val_reg = byt_gpio_reg(vg, offset, BYT_VAL_REG); + void __iomem *db_reg = byt_gpio_reg(vg, offset, BYT_DEBOUNCE_REG); unsigned long flags; u32 conf, pull, val, debounce; u16 arg = 0; @@ -1128,7 +1129,7 @@ static int byt_pin_config_get(struct pinctrl_dev *pctl_dev, unsigned int offset, return -EINVAL; raw_spin_lock_irqsave(&vg->lock, flags); - debounce = readl(byt_gpio_reg(vg, offset, BYT_DEBOUNCE_REG)); + debounce = readl(db_reg); raw_spin_unlock_irqrestore(&vg->lock, flags); switch (debounce & BYT_DEBOUNCE_PULSE_MASK) { @@ -1176,6 +1177,7 @@ static int byt_pin_config_set(struct pinctrl_dev *pctl_dev, unsigned int param, arg; void __iomem *conf_reg = byt_gpio_reg(vg, offset, BYT_CONF0_REG); void __iomem *val_reg = byt_gpio_reg(vg, offset, BYT_VAL_REG); + void __iomem *db_reg = byt_gpio_reg(vg, offset, BYT_DEBOUNCE_REG); unsigned long flags; u32 conf, val, debounce; int i, ret = 0; @@ -1238,36 +1240,40 @@ static int byt_pin_config_set(struct pinctrl_dev *pctl_dev, break; case PIN_CONFIG_INPUT_DEBOUNCE: - debounce = readl(byt_gpio_reg(vg, offset, - BYT_DEBOUNCE_REG)); - conf &= ~BYT_DEBOUNCE_PULSE_MASK; + debounce = readl(db_reg); + debounce &= ~BYT_DEBOUNCE_PULSE_MASK; switch (arg) { + case 0: + conf &= BYT_DEBOUNCE_EN; + break; case 375: - conf |= BYT_DEBOUNCE_PULSE_375US; + debounce |= BYT_DEBOUNCE_PULSE_375US; break; case 750: - conf |= BYT_DEBOUNCE_PULSE_750US; + debounce |= BYT_DEBOUNCE_PULSE_750US; break; case 1500: - conf |= BYT_DEBOUNCE_PULSE_1500US; + debounce |= BYT_DEBOUNCE_PULSE_1500US; break; case 3000: - conf |= BYT_DEBOUNCE_PULSE_3MS; + debounce |= BYT_DEBOUNCE_PULSE_3MS; break; case 6000: - conf |= BYT_DEBOUNCE_PULSE_6MS; + debounce |= BYT_DEBOUNCE_PULSE_6MS; break; case 12000: - conf |= BYT_DEBOUNCE_PULSE_12MS; + debounce |= BYT_DEBOUNCE_PULSE_12MS; break; case 24000: - conf |= BYT_DEBOUNCE_PULSE_24MS; + debounce |= BYT_DEBOUNCE_PULSE_24MS; break; default: ret = -EINVAL; } + if (!ret) + writel(debounce, db_reg); break; default: ret = -ENOTSUPP; From 143a9ad4e68cc5c210e6e99e910d6b77cc8a9ec5 Mon Sep 17 00:00:00 2001 From: Yasuaki Ishimatsu Date: Tue, 24 Jan 2017 15:17:45 -0800 Subject: [PATCH 060/119] memory_hotplug: make zone_can_shift() return a boolean value commit 8a1f780e7f28c7c1d640118242cf68d528c456cd upstream. online_{kernel|movable} is used to change the memory zone to ZONE_{NORMAL|MOVABLE} and online the memory. To check that memory zone can be changed, zone_can_shift() is used. Currently the function returns minus integer value, plus integer value and 0. When the function returns minus or plus integer value, it means that the memory zone can be changed to ZONE_{NORNAL|MOVABLE}. But when the function returns 0, there are two meanings. One of the meanings is that the memory zone does not need to be changed. For example, when memory is in ZONE_NORMAL and onlined by online_kernel the memory zone does not need to be changed. Another meaning is that the memory zone cannot be changed. When memory is in ZONE_NORMAL and onlined by online_movable, the memory zone may not be changed to ZONE_MOVALBE due to memory online limitation(see Documentation/memory-hotplug.txt). In this case, memory must not be onlined. The patch changes the return type of zone_can_shift() so that memory online operation fails when memory zone cannot be changed as follows: Before applying patch: # grep -A 35 "Node 2" /proc/zoneinfo Node 2, zone Normal node_scanned 0 spanned 8388608 present 7864320 managed 7864320 # echo online_movable > memory4097/state # grep -A 35 "Node 2" /proc/zoneinfo Node 2, zone Normal node_scanned 0 spanned 8388608 present 8388608 managed 8388608 online_movable operation succeeded. But memory is onlined as ZONE_NORMAL, not ZONE_MOVABLE. After applying patch: # grep -A 35 "Node 2" /proc/zoneinfo Node 2, zone Normal node_scanned 0 spanned 8388608 present 7864320 managed 7864320 # echo online_movable > memory4097/state bash: echo: write error: Invalid argument # grep -A 35 "Node 2" /proc/zoneinfo Node 2, zone Normal node_scanned 0 spanned 8388608 present 7864320 managed 7864320 online_movable operation failed because of failure of changing the memory zone from ZONE_NORMAL to ZONE_MOVABLE Fixes: df429ac03936 ("memory-hotplug: more general validation of zone during online") Link: http://lkml.kernel.org/r/2f9c3837-33d7-b6e5-59c0-6ca4372b2d84@gmail.com Signed-off-by: Yasuaki Ishimatsu Reviewed-by: Reza Arbab Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- drivers/base/memory.c | 4 ++-- include/linux/memory_hotplug.h | 4 ++-- mm/memory_hotplug.c | 28 +++++++++++++++++----------- 3 files changed, 21 insertions(+), 15 deletions(-) diff --git a/drivers/base/memory.c b/drivers/base/memory.c index 62c63c0c5c22..e7f86a8887d2 100644 --- a/drivers/base/memory.c +++ b/drivers/base/memory.c @@ -410,14 +410,14 @@ static ssize_t show_valid_zones(struct device *dev, sprintf(buf, "%s", zone->name); /* MMOP_ONLINE_KERNEL */ - zone_shift = zone_can_shift(start_pfn, nr_pages, ZONE_NORMAL); + zone_can_shift(start_pfn, nr_pages, ZONE_NORMAL, &zone_shift); if (zone_shift) { strcat(buf, " "); strcat(buf, (zone + zone_shift)->name); } /* MMOP_ONLINE_MOVABLE */ - zone_shift = zone_can_shift(start_pfn, nr_pages, ZONE_MOVABLE); + zone_can_shift(start_pfn, nr_pages, ZONE_MOVABLE, &zone_shift); if (zone_shift) { strcat(buf, " "); strcat(buf, (zone + zone_shift)->name); diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h index 01033fadea47..c1784c0b4f35 100644 --- a/include/linux/memory_hotplug.h +++ b/include/linux/memory_hotplug.h @@ -284,7 +284,7 @@ extern void sparse_remove_one_section(struct zone *zone, struct mem_section *ms, unsigned long map_offset); extern struct page *sparse_decode_mem_map(unsigned long coded_mem_map, unsigned long pnum); -extern int zone_can_shift(unsigned long pfn, unsigned long nr_pages, - enum zone_type target); +extern bool zone_can_shift(unsigned long pfn, unsigned long nr_pages, + enum zone_type target, int *zone_shift); #endif /* __LINUX_MEMORY_HOTPLUG_H */ diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index cad4b9125695..c3a8141ac788 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -1033,36 +1033,39 @@ static void node_states_set_node(int node, struct memory_notify *arg) node_set_state(node, N_MEMORY); } -int zone_can_shift(unsigned long pfn, unsigned long nr_pages, - enum zone_type target) +bool zone_can_shift(unsigned long pfn, unsigned long nr_pages, + enum zone_type target, int *zone_shift) { struct zone *zone = page_zone(pfn_to_page(pfn)); enum zone_type idx = zone_idx(zone); int i; + *zone_shift = 0; + if (idx < target) { /* pages must be at end of current zone */ if (pfn + nr_pages != zone_end_pfn(zone)) - return 0; + return false; /* no zones in use between current zone and target */ for (i = idx + 1; i < target; i++) if (zone_is_initialized(zone - idx + i)) - return 0; + return false; } if (target < idx) { /* pages must be at beginning of current zone */ if (pfn != zone->zone_start_pfn) - return 0; + return false; /* no zones in use between current zone and target */ for (i = target + 1; i < idx; i++) if (zone_is_initialized(zone - idx + i)) - return 0; + return false; } - return target - idx; + *zone_shift = target - idx; + return true; } /* Must be protected by mem_hotplug_begin() */ @@ -1089,10 +1092,13 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ !can_online_high_movable(zone)) return -EINVAL; - if (online_type == MMOP_ONLINE_KERNEL) - zone_shift = zone_can_shift(pfn, nr_pages, ZONE_NORMAL); - else if (online_type == MMOP_ONLINE_MOVABLE) - zone_shift = zone_can_shift(pfn, nr_pages, ZONE_MOVABLE); + if (online_type == MMOP_ONLINE_KERNEL) { + if (!zone_can_shift(pfn, nr_pages, ZONE_NORMAL, &zone_shift)) + return -EINVAL; + } else if (online_type == MMOP_ONLINE_MOVABLE) { + if (!zone_can_shift(pfn, nr_pages, ZONE_MOVABLE, &zone_shift)) + return -EINVAL; + } zone = move_pfn_range(zone_shift, pfn, pfn + nr_pages); if (!zone) From 776050a9b55e17b72b5684794b9580d72e920e17 Mon Sep 17 00:00:00 2001 From: Robin Murphy Date: Tue, 10 Jan 2017 17:51:17 +0000 Subject: [PATCH 061/119] virtio_mmio: Set DMA masks appropriately commit f7f6634d23830ff74335734fbdb28ea109c1f349 upstream. Once DMA API usage is enabled, it becomes apparent that virtio-mmio is inadvertently relying on the default 32-bit DMA mask, which leads to problems like rapidly exhausting SWIOTLB bounce buffers. Ensure that we set the appropriate 64-bit DMA mask whenever possible, with the coherent mask suitably limited for the legacy vring as per a0be1db4304f ("virtio_pci: Limit DMA mask to 44 bits for legacy virtio devices"). Cc: Andy Lutomirski Cc: Michael S. Tsirkin Reported-by: Jean-Philippe Brucker Fixes: b42111382f0e ("virtio_mmio: Use the DMA API if enabled") Signed-off-by: Robin Murphy Signed-off-by: Michael S. Tsirkin Signed-off-by: Greg Kroah-Hartman --- drivers/virtio/virtio_mmio.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) diff --git a/drivers/virtio/virtio_mmio.c b/drivers/virtio/virtio_mmio.c index 48bfea91dbca..50840984fbfa 100644 --- a/drivers/virtio/virtio_mmio.c +++ b/drivers/virtio/virtio_mmio.c @@ -59,6 +59,7 @@ #define pr_fmt(fmt) "virtio-mmio: " fmt #include +#include #include #include #include @@ -497,6 +498,7 @@ static int virtio_mmio_probe(struct platform_device *pdev) struct virtio_mmio_device *vm_dev; struct resource *mem; unsigned long magic; + int rc; mem = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (!mem) @@ -545,9 +547,25 @@ static int virtio_mmio_probe(struct platform_device *pdev) } vm_dev->vdev.id.vendor = readl(vm_dev->base + VIRTIO_MMIO_VENDOR_ID); - if (vm_dev->version == 1) + if (vm_dev->version == 1) { writel(PAGE_SIZE, vm_dev->base + VIRTIO_MMIO_GUEST_PAGE_SIZE); + rc = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64)); + /* + * In the legacy case, ensure our coherently-allocated virtio + * ring will be at an address expressable as a 32-bit PFN. + */ + if (!rc) + dma_set_coherent_mask(&pdev->dev, + DMA_BIT_MASK(32 + PAGE_SHIFT)); + } else { + rc = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64)); + } + if (rc) + rc = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32)); + if (rc) + dev_warn(&pdev->dev, "Failed to enable 64-bit or 32-bit DMA. Trying to continue, but this might not work.\n"); + platform_set_drvdata(pdev, vm_dev); return register_virtio_device(&vm_dev->vdev); From adb260d15134f3aabd6e1bf39ee7f4ea28385c2d Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 7 Jan 2017 09:33:34 +0300 Subject: [PATCH 062/119] platform/x86: mlx-platform: free first dev on error commit 63d762b88cb5510f2bfdb5112ced18cde867ae61 upstream. There is an off-by-one error so we don't unregister priv->pdev_mux[0]. Also it's slightly simpler as a while loop instead of a for loop. Fixes: 58cbbee2391c ("x86/platform/mellanox: Introduce support for Mellanox systems platform") Signed-off-by: Dan Carpenter Acked-by: Vadim Pasternak Signed-off-by: Andy Shevchenko Signed-off-by: Greg Kroah-Hartman --- arch/x86/platform/mellanox/mlx-platform.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/platform/mellanox/mlx-platform.c b/arch/x86/platform/mellanox/mlx-platform.c index 7dcfcca97399..c0355d789fce 100644 --- a/arch/x86/platform/mellanox/mlx-platform.c +++ b/arch/x86/platform/mellanox/mlx-platform.c @@ -233,7 +233,7 @@ static int __init mlxplat_init(void) return 0; fail_platform_mux_register: - for (i--; i > 0 ; i--) + while (--i >= 0) platform_device_unregister(priv->pdev_mux[i]); platform_device_unregister(priv->pdev_i2c); fail_alloc: From 98185d4b18a14dad0b95efd85225960b5435727d Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Thu, 19 Jan 2017 18:39:40 +0200 Subject: [PATCH 063/119] platform/x86: intel_mid_powerbtn: Set IRQ_ONESHOT commit 5a00b6c2438460b870a451f14593fc40d3c7edf6 upstream. The commit 1c6c69525b40 ("genirq: Reject bogus threaded irq requests") starts refusing misconfigured interrupt handlers. This makes intel_mid_powerbtn not working anymore. Add a mandatory flag to a threaded IRQ request in the driver. Fixes: 1c6c69525b40 ("genirq: Reject bogus threaded irq requests") Signed-off-by: Andy Shevchenko Signed-off-by: Greg Kroah-Hartman --- drivers/platform/x86/intel_mid_powerbtn.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/platform/x86/intel_mid_powerbtn.c b/drivers/platform/x86/intel_mid_powerbtn.c index 1fc0de870ff8..361770568ad0 100644 --- a/drivers/platform/x86/intel_mid_powerbtn.c +++ b/drivers/platform/x86/intel_mid_powerbtn.c @@ -77,7 +77,7 @@ static int mfld_pb_probe(struct platform_device *pdev) input_set_capability(input, EV_KEY, KEY_POWER); - error = request_threaded_irq(irq, NULL, mfld_pb_isr, 0, + error = request_threaded_irq(irq, NULL, mfld_pb_isr, IRQF_ONESHOT, DRIVER_NAME, input); if (error) { dev_err(&pdev->dev, "Unable to request irq %d for mfld power" From f5f415c13209c5a7a71da85462815288321483df Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Tue, 24 Jan 2017 15:18:10 -0800 Subject: [PATCH 064/119] mm, memcg: do not retry precharge charges commit 3674534b775354516e5c148ea48f51d4d1909a78 upstream. When memory.move_charge_at_immigrate is enabled and precharges are depleted during move, mem_cgroup_move_charge_pte_range() will attempt to increase the size of the precharge. Prevent precharges from ever looping by setting __GFP_NORETRY. This was probably the intention of the GFP_KERNEL & ~__GFP_NORETRY, which is pointless as written. Fixes: 0029e19ebf84 ("mm: memcontrol: remove explicit OOM parameter in charge path") Link: http://lkml.kernel.org/r/alpine.DEB.2.10.1701130208510.69402@chino.kir.corp.google.com Signed-off-by: David Rientjes Acked-by: Michal Hocko Cc: Johannes Weiner Cc: Vladimir Davydov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- mm/memcontrol.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index d536a9daa511..4c6ade54d833 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -4360,9 +4360,9 @@ static int mem_cgroup_do_precharge(unsigned long count) return ret; } - /* Try charges one by one with reclaim */ + /* Try charges one by one with reclaim, but do not retry */ while (count--) { - ret = try_charge(mc.to, GFP_KERNEL & ~__GFP_NORETRY, 1); + ret = try_charge(mc.to, GFP_KERNEL | __GFP_NORETRY, 1); if (ret) return ret; mc.precharge++; From 922813f4d66fb317e8602d058d03a1619af1ffd0 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 11 Jan 2017 21:09:50 +0100 Subject: [PATCH 065/119] perf/core: Fix concurrent sys_perf_event_open() vs. 'move_group' race commit 321027c1fe77f892f4ea07846aeae08cefbbb290 upstream. Di Shen reported a race between two concurrent sys_perf_event_open() calls where both try and move the same pre-existing software group into a hardware context. The problem is exactly that described in commit: f63a8daa5812 ("perf: Fix event->ctx locking") ... where, while we wait for a ctx->mutex acquisition, the event->ctx relation can have changed under us. That very same commit failed to recognise sys_perf_event_context() as an external access vector to the events and thereby didn't apply the established locking rules correctly. So while one sys_perf_event_open() call is stuck waiting on mutex_lock_double(), the other (which owns said locks) moves the group about. So by the time the former sys_perf_event_open() acquires the locks, the context we've acquired is stale (and possibly dead). Apply the established locking rules as per perf_event_ctx_lock_nested() to the mutex_lock_double() for the 'move_group' case. This obviously means we need to validate state after we acquire the locks. Reported-by: Di Shen (Keen Lab) Tested-by: John Dias Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Kees Cook Cc: Linus Torvalds Cc: Min Chong Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Fixes: f63a8daa5812 ("perf: Fix event->ctx locking") Link: http://lkml.kernel.org/r/20170106131444.GZ3174@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar Signed-off-by: Greg Kroah-Hartman --- kernel/events/core.c | 58 +++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 54 insertions(+), 4 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index 02c8421f8c01..e5a8839e7076 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -9503,6 +9503,37 @@ static int perf_event_set_clock(struct perf_event *event, clockid_t clk_id) return 0; } +/* + * Variation on perf_event_ctx_lock_nested(), except we take two context + * mutexes. + */ +static struct perf_event_context * +__perf_event_ctx_lock_double(struct perf_event *group_leader, + struct perf_event_context *ctx) +{ + struct perf_event_context *gctx; + +again: + rcu_read_lock(); + gctx = READ_ONCE(group_leader->ctx); + if (!atomic_inc_not_zero(&gctx->refcount)) { + rcu_read_unlock(); + goto again; + } + rcu_read_unlock(); + + mutex_lock_double(&gctx->mutex, &ctx->mutex); + + if (group_leader->ctx != gctx) { + mutex_unlock(&ctx->mutex); + mutex_unlock(&gctx->mutex); + put_ctx(gctx); + goto again; + } + + return gctx; +} + /** * sys_perf_event_open - open a performance event, associate it to a task/cpu * @@ -9746,12 +9777,31 @@ SYSCALL_DEFINE5(perf_event_open, } if (move_group) { - gctx = group_leader->ctx; - mutex_lock_double(&gctx->mutex, &ctx->mutex); + gctx = __perf_event_ctx_lock_double(group_leader, ctx); + if (gctx->task == TASK_TOMBSTONE) { err = -ESRCH; goto err_locked; } + + /* + * Check if we raced against another sys_perf_event_open() call + * moving the software group underneath us. + */ + if (!(group_leader->group_caps & PERF_EV_CAP_SOFTWARE)) { + /* + * If someone moved the group out from under us, check + * if this new event wound up on the same ctx, if so + * its the regular !move_group case, otherwise fail. + */ + if (gctx != ctx) { + err = -EINVAL; + goto err_locked; + } else { + perf_event_ctx_unlock(group_leader, gctx); + move_group = 0; + } + } } else { mutex_lock(&ctx->mutex); } @@ -9853,7 +9903,7 @@ SYSCALL_DEFINE5(perf_event_open, perf_unpin_context(ctx); if (move_group) - mutex_unlock(&gctx->mutex); + perf_event_ctx_unlock(group_leader, gctx); mutex_unlock(&ctx->mutex); if (task) { @@ -9879,7 +9929,7 @@ SYSCALL_DEFINE5(perf_event_open, err_locked: if (move_group) - mutex_unlock(&gctx->mutex); + perf_event_ctx_unlock(group_leader, gctx); mutex_unlock(&ctx->mutex); /* err_file: */ fput(event_file); From b59dd202f231431c256b56de129e979d9563612f Mon Sep 17 00:00:00 2001 From: Francisco Jerez Date: Thu, 12 Jan 2017 12:44:54 +0200 Subject: [PATCH 066/119] drm/i915: Remove WaDisableLSQCROPERFforOCL KBL workaround. commit 4fc020d864647ea3ae8cb8f17d63e48e87ebd0bf upstream. The WaDisableLSQCROPERFforOCL workaround has the side effect of disabling an L3SQ optimization that has huge performance implications and is unlikely to be necessary for the correct functioning of usual graphic workloads. Userspace is free to re-enable the workaround on demand, and is generally in a better position to determine whether the workaround is necessary than the DRM is (e.g. only during the execution of compute kernels that rely on both L3 fences and HDC R/W requests). The same workaround seems to apply to BDW (at least to production stepping G1) and SKL as well (the internal workaround database claims that it does for all steppings, while the BSpec workaround table only mentions pre-production steppings), but the DRM doesn't do anything beyond whitelisting the L3SQCREG4 register so userspace can enable it when it sees fit. Do the same on KBL platforms. Improves performance of the GFXBench4 gl_manhattan31 benchmark by 60%, and gl_4 (AKA car chase) by 14% on a KBL GT2 running Mesa master -- This is followed by a regression of 35% and 10% respectively for the same benchmarks and platform caused by my recent patch series switching userspace to use the dataport constant cache instead of the sampler to implement uniform pull constant loads, which caused us to hit more heavily the L3 cache (and on platforms other than KBL had the opposite effect of improving performance of the same two benchmarks). The overall effect on KBL of this change combined with the recent userspace change is respectively 4.6% and 2.6%. SynMark2 OglShMapPcf was affected by the constant cache changes (though it improved as it did on other platforms rather than regressing), but is not significantly affected by this patch (with statistical significance of 5% and sample size 20). v2: Drop some more code to avoid unused variable warning. Fixes: 738fa1b3123f ("drm/i915/kbl: Add WaDisableLSQCROPERFforOCL") Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99256 Signed-off-by: Francisco Jerez Cc: Matthew Auld Cc: Eero Tamminen Cc: Jani Nikula Cc: Mika Kuoppala Cc: beignet@lists.freedesktop.org Reviewed-by: Mika Kuoppala [Removed double Fixes tag] Signed-off-by: Mika Kuoppala Link: http://patchwork.freedesktop.org/patch/msgid/1484217894-20505-1-git-send-email-mika.kuoppala@intel.com (cherry picked from commit 8726f2faa371514fba2f594d799db95203dfeee0) Signed-off-by: Jani Nikula [ Francisco Jerez: Rebase on v4.9 branch. ] Signed-off-by: Francisco Jerez Signed-off-by: Greg Kroah-Hartman --- drivers/gpu/drm/i915/intel_lrc.c | 3 +-- drivers/gpu/drm/i915/intel_ringbuffer.c | 8 -------- 2 files changed, 1 insertion(+), 10 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 0adb879833ff..67db1577ee49 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -858,8 +858,7 @@ static inline int gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, * this batch updates GEN8_L3SQCREG4 with default value we need to * set this bit here to retain the WA during flush. */ - if (IS_SKL_REVID(dev_priv, 0, SKL_REVID_E0) || - IS_KBL_REVID(dev_priv, 0, KBL_REVID_E0)) + if (IS_SKL_REVID(dev_priv, 0, SKL_REVID_E0)) l3sqc4_flush |= GEN8_LQSC_RO_PERF_DIS; wa_ctx_emit(batch, index, (MI_STORE_REGISTER_MEM_GEN8 | diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index ed9955dce156..8babfe0ce4e3 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -1153,14 +1153,6 @@ static int kbl_init_workarounds(struct intel_engine_cs *engine) WA_SET_BIT_MASKED(HDC_CHICKEN0, HDC_FENCE_DEST_SLM_DISABLE); - /* GEN8_L3SQCREG4 has a dependency with WA batch so any new changes - * involving this register should also be added to WA batch as required. - */ - if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_E0)) - /* WaDisableLSQCROPERFforOCL:kbl */ - I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) | - GEN8_LQSC_RO_PERF_DIS); - /* WaToEnableHwFixForPushConstHWBug:kbl */ if (IS_KBL_REVID(dev_priv, KBL_REVID_C0, REVID_FOREVER)) WA_SET_BIT_MASKED(COMMON_SLICE_CHICKEN2, From fd2ffe57dda03cb070204f53864ecdfd002aa2e3 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Wed, 1 Feb 2017 08:33:31 +0100 Subject: [PATCH 067/119] Linux 4.9.7 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index ef95231d1625..da704d903321 100644 --- a/Makefile +++ b/Makefile @@ -1,6 +1,6 @@ VERSION = 4 PATCHLEVEL = 9 -SUBLEVEL = 6 +SUBLEVEL = 7 EXTRAVERSION = NAME = Roaring Lionus From a37f2311e63c8989675e297cdc9300cbec6e2217 Mon Sep 17 00:00:00 2001 From: hayeswang Date: Wed, 11 Jan 2017 16:25:34 +0800 Subject: [PATCH 068/119] r8152: fix the sw rx checksum is unavailable [ Upstream commit 19c0f40d4fca3a47b8f784a627f0467f0138ccc8 ] Fix the hw rx checksum is always enabled, and the user couldn't switch it to sw rx checksum. Note that the RTL_VER_01 only support sw rx checksum only. Besides, the hw rx checksum for RTL_VER_02 is disabled after commit b9a321b48af4 ("r8152: Fix broken RX checksums."). Re-enable it. Signed-off-by: Hayes Wang Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/usb/r8152.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c index 4b5cb162442b..be4c86faba4f 100644 --- a/drivers/net/usb/r8152.c +++ b/drivers/net/usb/r8152.c @@ -1730,7 +1730,7 @@ static u8 r8152_rx_csum(struct r8152 *tp, struct rx_desc *rx_desc) u8 checksum = CHECKSUM_NONE; u32 opts2, opts3; - if (tp->version == RTL_VER_01 || tp->version == RTL_VER_02) + if (!(tp->netdev->features & NETIF_F_RXCSUM)) goto return_result; opts2 = le32_to_cpu(rx_desc->opts2); @@ -4358,6 +4358,11 @@ static int rtl8152_probe(struct usb_interface *intf, NETIF_F_HIGHDMA | NETIF_F_FRAGLIST | NETIF_F_IPV6_CSUM | NETIF_F_TSO6; + if (tp->version == RTL_VER_01) { + netdev->features &= ~NETIF_F_RXCSUM; + netdev->hw_features &= ~NETIF_F_RXCSUM; + } + netdev->ethtool_ops = &ops; netif_set_gso_max_size(netdev, RTL_LIMITED_TSO_SIZE); From 5b3df4401064ef94cddac976aba1474a7050fa5e Mon Sep 17 00:00:00 2001 From: stephen hemminger Date: Wed, 11 Jan 2017 09:16:32 -0800 Subject: [PATCH 069/119] netvsc: add rcu_read locking to netvsc callback [ Upstream commit 0719e72ccb801829a3d735d187ca8417f0930459 ] The receive callback (in tasklet context) is using RCU to get reference to associated VF network device but this is not safe. RCU read lock needs to be held. Found by running with full lockdep debugging enabled. Fixes: f207c10d9823 ("hv_netvsc: use RCU to protect vf_netdev") Signed-off-by: Stephen Hemminger Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/hyperv/netvsc_drv.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/hyperv/netvsc_drv.c b/drivers/net/hyperv/netvsc_drv.c index c9140c3aeb67..ff038e507fd6 100644 --- a/drivers/net/hyperv/netvsc_drv.c +++ b/drivers/net/hyperv/netvsc_drv.c @@ -659,6 +659,7 @@ int netvsc_recv_callback(struct hv_device *device_obj, * policy filters on the host). Deliver these via the VF * interface in the guest. */ + rcu_read_lock(); vf_netdev = rcu_dereference(net_device_ctx->vf_netdev); if (vf_netdev && (vf_netdev->flags & IFF_UP)) net = vf_netdev; @@ -667,6 +668,7 @@ int netvsc_recv_callback(struct hv_device *device_obj, skb = netvsc_alloc_recv_skb(net, packet, csum_info, *data, vlan_tci); if (unlikely(!skb)) { ++net->stats.rx_dropped; + rcu_read_unlock(); return NVSP_STAT_FAIL; } @@ -696,6 +698,7 @@ int netvsc_recv_callback(struct hv_device *device_obj, * TODO - use NAPI? */ netif_rx(skb); + rcu_read_unlock(); return 0; } From 7c249f3306b6f42f7a0a9d9cbd0455035f8a067c Mon Sep 17 00:00:00 2001 From: Arkadi Sharshevsky Date: Thu, 12 Jan 2017 09:10:37 +0100 Subject: [PATCH 070/119] mlxsw: spectrum: Fix memory leak at skb reallocation [ Upstream commit 36bf38d158d3482119b3e159c0619b3c1539b508 ] During transmission the skb is checked for headroom in order to add vendor specific header. In case the skb needs to be re-allocated, skb_realloc_headroom() is called to make a private copy of the original, but doesn't release it. Current code assumes that the original skb is released during reallocation and only releases it at the error path which causes a memory leak. Fix this by adding the original skb release to the main path. Fixes: 56ade8fe3fe1 ("mlxsw: spectrum: Add initial support for Spectrum ASIC") Signed-off-by: Arkadi Sharshevsky Reviewed-by: Ido Schimmel Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/mellanox/mlxsw/spectrum.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index dda5761e91bc..f902c4d3de99 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -684,6 +684,7 @@ static netdev_tx_t mlxsw_sp_port_xmit(struct sk_buff *skb, dev_kfree_skb_any(skb_orig); return NETDEV_TX_OK; } + dev_consume_skb_any(skb_orig); } if (eth_skb_pad(skb)) { From 4ec59d1fe470bad37dc0d388f65c150ccce8ea11 Mon Sep 17 00:00:00 2001 From: Arkadi Sharshevsky Date: Thu, 12 Jan 2017 09:10:38 +0100 Subject: [PATCH 071/119] mlxsw: switchx2: Fix memory leak at skb reallocation [ Upstream commit 400fc0106dd8c27ed84781c929c1a184785b9c79 ] During transmission the skb is checked for headroom in order to add vendor specific header. In case the skb needs to be re-allocated, skb_realloc_headroom() is called to make a private copy of the original, but doesn't release it. Current code assumes that the original skb is released during reallocation and only releases it at the error path which causes a memory leak. Fix this by adding the original skb release to the main path. Fixes: d003462a50de ("mlxsw: Simplify mlxsw_sx_port_xmit function") Signed-off-by: Arkadi Sharshevsky Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/mellanox/mlxsw/switchx2.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/ethernet/mellanox/mlxsw/switchx2.c b/drivers/net/ethernet/mellanox/mlxsw/switchx2.c index 92bda8703f87..d548f0a55174 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/switchx2.c +++ b/drivers/net/ethernet/mellanox/mlxsw/switchx2.c @@ -314,6 +314,7 @@ static netdev_tx_t mlxsw_sx_port_xmit(struct sk_buff *skb, dev_kfree_skb_any(skb_orig); return NETDEV_TX_OK; } + dev_consume_skb_any(skb_orig); } mlxsw_sx_txhdr_construct(skb, &tx_info); /* TX header is consumed by HW on the way so we shouldn't count its From ec1aa8d495326435d956af325391b775038daa08 Mon Sep 17 00:00:00 2001 From: Elad Raz Date: Thu, 12 Jan 2017 09:10:39 +0100 Subject: [PATCH 072/119] mlxsw: pci: Fix EQE structure definition [ Upstream commit 28e46a0f2e03ab4ed0e23cace1ea89a68c8c115b ] The event_data starts from address 0x00-0x0C and not from 0x08-0x014. This leads to duplication with other fields in the Event Queue Element such as sub-type, cqn and owner. Fixes: eda6500a987a0 ("mlxsw: Add PCI bus implementation") Signed-off-by: Elad Raz Signed-off-by: Jiri Pirko Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/mellanox/mlxsw/pci.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlxsw/pci.h b/drivers/net/ethernet/mellanox/mlxsw/pci.h index d942a3e6fa41..846fd4df7dab 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/pci.h +++ b/drivers/net/ethernet/mellanox/mlxsw/pci.h @@ -211,21 +211,21 @@ MLXSW_ITEM32(pci, eqe, owner, 0x0C, 0, 1); /* pci_eqe_cmd_token * Command completion event - token */ -MLXSW_ITEM32(pci, eqe, cmd_token, 0x08, 16, 16); +MLXSW_ITEM32(pci, eqe, cmd_token, 0x00, 16, 16); /* pci_eqe_cmd_status * Command completion event - status */ -MLXSW_ITEM32(pci, eqe, cmd_status, 0x08, 0, 8); +MLXSW_ITEM32(pci, eqe, cmd_status, 0x00, 0, 8); /* pci_eqe_cmd_out_param_h * Command completion event - output parameter - higher part */ -MLXSW_ITEM32(pci, eqe, cmd_out_param_h, 0x0C, 0, 32); +MLXSW_ITEM32(pci, eqe, cmd_out_param_h, 0x04, 0, 32); /* pci_eqe_cmd_out_param_l * Command completion event - output parameter - lower part */ -MLXSW_ITEM32(pci, eqe, cmd_out_param_l, 0x10, 0, 32); +MLXSW_ITEM32(pci, eqe, cmd_out_param_l, 0x08, 0, 32); #endif From 6980c52c4efb951c972409ebd146cfc348144918 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Wed, 11 Jan 2017 14:29:54 -0800 Subject: [PATCH 073/119] net: lwtunnel: Handle lwtunnel_fill_encap failure [ Upstream commit ea7a80858f57d8878b1499ea0f1b8a635cc48de7 ] Handle failure in lwtunnel_fill_encap adding attributes to skb. Fixes: 571e722676fe ("ipv4: support for fib route lwtunnel encap attributes") Fixes: 19e42e451506 ("ipv6: support for fib route lwtunnel encap attributes") Signed-off-by: David Ahern Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/ipv4/fib_semantics.c | 11 +++++++---- net/ipv6/route.c | 3 ++- 2 files changed, 9 insertions(+), 5 deletions(-) diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c index a8508b79b406..6a4068031aaa 100644 --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -1278,8 +1278,9 @@ int fib_dump_info(struct sk_buff *skb, u32 portid, u32 seq, int event, nla_put_u32(skb, RTA_FLOW, fi->fib_nh[0].nh_tclassid)) goto nla_put_failure; #endif - if (fi->fib_nh->nh_lwtstate) - lwtunnel_fill_encap(skb, fi->fib_nh->nh_lwtstate); + if (fi->fib_nh->nh_lwtstate && + lwtunnel_fill_encap(skb, fi->fib_nh->nh_lwtstate) < 0) + goto nla_put_failure; } #ifdef CONFIG_IP_ROUTE_MULTIPATH if (fi->fib_nhs > 1) { @@ -1315,8 +1316,10 @@ int fib_dump_info(struct sk_buff *skb, u32 portid, u32 seq, int event, nla_put_u32(skb, RTA_FLOW, nh->nh_tclassid)) goto nla_put_failure; #endif - if (nh->nh_lwtstate) - lwtunnel_fill_encap(skb, nh->nh_lwtstate); + if (nh->nh_lwtstate && + lwtunnel_fill_encap(skb, nh->nh_lwtstate) < 0) + goto nla_put_failure; + /* length of rtnetlink header + attributes */ rtnh->rtnh_len = nlmsg_get_pos(skb) - (void *) rtnh; } endfor_nexthops(fi); diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 1b57e11e6e0d..acd8023763cb 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -3306,7 +3306,8 @@ static int rt6_fill_node(struct net *net, if (nla_put_u8(skb, RTA_PREF, IPV6_EXTRACT_PREF(rt->rt6i_flags))) goto nla_put_failure; - lwtunnel_fill_encap(skb, rt->dst.lwtstate); + if (lwtunnel_fill_encap(skb, rt->dst.lwtstate) < 0) + goto nla_put_failure; nlmsg_end(skb, nlh); return 0; From 958bb1bdc2c272fb63f31510d00579f59acb7a06 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Wed, 11 Jan 2017 15:42:17 -0800 Subject: [PATCH 074/119] net: ipv4: fix table id in getroute response [ Upstream commit 8a430ed50bb1b19ca14a46661f3b1b35f2fb5c39 ] rtm_table is an 8-bit field while table ids are allowed up to u32. Commit 709772e6e065 ("net: Fix routing tables with id > 255 for legacy software") added the preference to set rtm_table in dumps to RT_TABLE_COMPAT if the table id is > 255. The table id returned on get route requests should do the same. Fixes: c36ba6603a11 ("net: Allow user to get table id from route lookup") Signed-off-by: David Ahern Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/ipv4/route.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 8197b06d9aaa..d851cae27dac 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -2440,7 +2440,7 @@ static int rt_fill_info(struct net *net, __be32 dst, __be32 src, u32 table_id, r->rtm_dst_len = 32; r->rtm_src_len = 0; r->rtm_tos = fl4->flowi4_tos; - r->rtm_table = table_id; + r->rtm_table = table_id < 256 ? table_id : RT_TABLE_COMPAT; if (nla_put_u32(skb, RTA_TABLE, table_id)) goto nla_put_failure; r->rtm_type = rt->rt_type; From b66b1f5ac33d20d12fb56b30a1dac8d79c8a0038 Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Thu, 12 Jan 2017 12:09:09 -0800 Subject: [PATCH 075/119] net: systemport: Decouple flow control from __bcm_sysport_tx_reclaim [ Upstream commit 148d3d021cf9724fcf189ce4e525a094bbf5ce89 ] The __bcm_sysport_tx_reclaim() function is used to reclaim transmit resources in different places within the driver. Most of them should not affect the state of the transit flow control. Introduce bcm_sysport_tx_clean() which cleans the ring, but does not re-enable flow control towards the networking stack, and make bcm_sysport_tx_reclaim() do the actual transmit queue flow control. Fixes: 80105befdb4b ("net: systemport: add Broadcom SYSTEMPORT Ethernet MAC driver") Signed-off-by: Florian Fainelli Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/broadcom/bcmsysport.c | 25 ++++++++++++++++------ 1 file changed, 18 insertions(+), 7 deletions(-) diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c index 25d1eb4933d0..be7ec5a76a54 100644 --- a/drivers/net/ethernet/broadcom/bcmsysport.c +++ b/drivers/net/ethernet/broadcom/bcmsysport.c @@ -710,11 +710,8 @@ static unsigned int __bcm_sysport_tx_reclaim(struct bcm_sysport_priv *priv, unsigned int c_index, last_c_index, last_tx_cn, num_tx_cbs; unsigned int pkts_compl = 0, bytes_compl = 0; struct bcm_sysport_cb *cb; - struct netdev_queue *txq; u32 hw_ind; - txq = netdev_get_tx_queue(ndev, ring->index); - /* Compute how many descriptors have been processed since last call */ hw_ind = tdma_readl(priv, TDMA_DESC_RING_PROD_CONS_INDEX(ring->index)); c_index = (hw_ind >> RING_CONS_INDEX_SHIFT) & RING_CONS_INDEX_MASK; @@ -745,9 +742,6 @@ static unsigned int __bcm_sysport_tx_reclaim(struct bcm_sysport_priv *priv, ring->c_index = c_index; - if (netif_tx_queue_stopped(txq) && pkts_compl) - netif_tx_wake_queue(txq); - netif_dbg(priv, tx_done, ndev, "ring=%d c_index=%d pkts_compl=%d, bytes_compl=%d\n", ring->index, ring->c_index, pkts_compl, bytes_compl); @@ -759,16 +753,33 @@ static unsigned int __bcm_sysport_tx_reclaim(struct bcm_sysport_priv *priv, static unsigned int bcm_sysport_tx_reclaim(struct bcm_sysport_priv *priv, struct bcm_sysport_tx_ring *ring) { + struct netdev_queue *txq; unsigned int released; unsigned long flags; + txq = netdev_get_tx_queue(priv->netdev, ring->index); + spin_lock_irqsave(&ring->lock, flags); released = __bcm_sysport_tx_reclaim(priv, ring); + if (released) + netif_tx_wake_queue(txq); + spin_unlock_irqrestore(&ring->lock, flags); return released; } +/* Locked version of the per-ring TX reclaim, but does not wake the queue */ +static void bcm_sysport_tx_clean(struct bcm_sysport_priv *priv, + struct bcm_sysport_tx_ring *ring) +{ + unsigned long flags; + + spin_lock_irqsave(&ring->lock, flags); + __bcm_sysport_tx_reclaim(priv, ring); + spin_unlock_irqrestore(&ring->lock, flags); +} + static int bcm_sysport_tx_poll(struct napi_struct *napi, int budget) { struct bcm_sysport_tx_ring *ring = @@ -1253,7 +1264,7 @@ static void bcm_sysport_fini_tx_ring(struct bcm_sysport_priv *priv, napi_disable(&ring->napi); netif_napi_del(&ring->napi); - bcm_sysport_tx_reclaim(priv, ring); + bcm_sysport_tx_clean(priv, ring); kfree(ring->cbs); ring->cbs = NULL; From 3524f64224bdef1c834e2952aaa72c175621e2e3 Mon Sep 17 00:00:00 2001 From: Shannon Nelson Date: Thu, 12 Jan 2017 14:24:58 -0800 Subject: [PATCH 076/119] tcp: fix tcp_fastopen unaligned access complaints on sparc [ Upstream commit 003c941057eaa868ca6fedd29a274c863167230d ] Fix up a data alignment issue on sparc by swapping the order of the cookie byte array field with the length field in struct tcp_fastopen_cookie, and making it a proper union to clean up the typecasting. This addresses log complaints like these: log_unaligned: 113 callbacks suppressed Kernel unaligned access at TPC[976490] tcp_try_fastopen+0x2d0/0x360 Kernel unaligned access at TPC[9764ac] tcp_try_fastopen+0x2ec/0x360 Kernel unaligned access at TPC[9764c8] tcp_try_fastopen+0x308/0x360 Kernel unaligned access at TPC[9764e4] tcp_try_fastopen+0x324/0x360 Kernel unaligned access at TPC[976490] tcp_try_fastopen+0x2d0/0x360 Cc: Eric Dumazet Signed-off-by: Shannon Nelson Acked-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- include/linux/tcp.h | 7 ++++++- net/ipv4/tcp_fastopen.c | 2 +- 2 files changed, 7 insertions(+), 2 deletions(-) diff --git a/include/linux/tcp.h b/include/linux/tcp.h index a17ae7b85218..647532b0eb03 100644 --- a/include/linux/tcp.h +++ b/include/linux/tcp.h @@ -62,8 +62,13 @@ static inline unsigned int tcp_optlen(const struct sk_buff *skb) /* TCP Fast Open Cookie as stored in memory */ struct tcp_fastopen_cookie { + union { + u8 val[TCP_FASTOPEN_COOKIE_MAX]; +#if IS_ENABLED(CONFIG_IPV6) + struct in6_addr addr; +#endif + }; s8 len; - u8 val[TCP_FASTOPEN_COOKIE_MAX]; bool exp; /* In RFC6994 experimental option format */ }; diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c index 4e777a3243f9..f51919535ca7 100644 --- a/net/ipv4/tcp_fastopen.c +++ b/net/ipv4/tcp_fastopen.c @@ -113,7 +113,7 @@ static bool tcp_fastopen_cookie_gen(struct request_sock *req, struct tcp_fastopen_cookie tmp; if (__tcp_fastopen_cookie_gen(&ip6h->saddr, &tmp)) { - struct in6_addr *buf = (struct in6_addr *) tmp.val; + struct in6_addr *buf = &tmp.addr; int i; for (i = 0; i < 4; i++) From 18767acb7b67b27ec4355ca623cbaa209edc8f13 Mon Sep 17 00:00:00 2001 From: Lance Richardson Date: Thu, 12 Jan 2017 19:33:18 -0500 Subject: [PATCH 077/119] openvswitch: maintain correct checksum state in conntrack actions [ Upstream commit 75f01a4c9cc291ff5cb28ca1216adb163b7a20ee ] When executing conntrack actions on skbuffs with checksum mode CHECKSUM_COMPLETE, the checksum must be updated to account for header pushes and pulls. Otherwise we get "hw csum failure" logs similar to this (ICMP packet received on geneve tunnel via ixgbe NIC): [ 405.740065] genev_sys_6081: hw csum failure [ 405.740106] CPU: 3 PID: 0 Comm: swapper/3 Tainted: G I 4.10.0-rc3+ #1 [ 405.740108] Call Trace: [ 405.740110] [ 405.740113] dump_stack+0x63/0x87 [ 405.740116] netdev_rx_csum_fault+0x3a/0x40 [ 405.740118] __skb_checksum_complete+0xcf/0xe0 [ 405.740120] nf_ip_checksum+0xc8/0xf0 [ 405.740124] icmp_error+0x1de/0x351 [nf_conntrack_ipv4] [ 405.740132] nf_conntrack_in+0xe1/0x550 [nf_conntrack] [ 405.740137] ? find_bucket.isra.2+0x62/0x70 [openvswitch] [ 405.740143] __ovs_ct_lookup+0x95/0x980 [openvswitch] [ 405.740145] ? netif_rx_internal+0x44/0x110 [ 405.740149] ovs_ct_execute+0x147/0x4b0 [openvswitch] [ 405.740153] do_execute_actions+0x22e/0xa70 [openvswitch] [ 405.740157] ovs_execute_actions+0x40/0x120 [openvswitch] [ 405.740161] ovs_dp_process_packet+0x84/0x120 [openvswitch] [ 405.740166] ovs_vport_receive+0x73/0xd0 [openvswitch] [ 405.740168] ? udp_rcv+0x1a/0x20 [ 405.740170] ? ip_local_deliver_finish+0x93/0x1e0 [ 405.740172] ? ip_local_deliver+0x6f/0xe0 [ 405.740174] ? ip_rcv_finish+0x3a0/0x3a0 [ 405.740176] ? ip_rcv_finish+0xdb/0x3a0 [ 405.740177] ? ip_rcv+0x2a7/0x400 [ 405.740180] ? __netif_receive_skb_core+0x970/0xa00 [ 405.740185] netdev_frame_hook+0xd3/0x160 [openvswitch] [ 405.740187] __netif_receive_skb_core+0x1dc/0xa00 [ 405.740194] ? ixgbe_clean_rx_irq+0x46d/0xa20 [ixgbe] [ 405.740197] __netif_receive_skb+0x18/0x60 [ 405.740199] netif_receive_skb_internal+0x40/0xb0 [ 405.740201] napi_gro_receive+0xcd/0x120 [ 405.740204] gro_cell_poll+0x57/0x80 [geneve] [ 405.740206] net_rx_action+0x260/0x3c0 [ 405.740209] __do_softirq+0xc9/0x28c [ 405.740211] irq_exit+0xd9/0xf0 [ 405.740213] do_IRQ+0x51/0xd0 [ 405.740215] common_interrupt+0x93/0x93 Fixes: 7f8a436eaa2c ("openvswitch: Add conntrack action") Signed-off-by: Lance Richardson Acked-by: Pravin B Shelar Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/openvswitch/conntrack.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/openvswitch/conntrack.c b/net/openvswitch/conntrack.c index fecefa2dc94e..eab210bb1ef0 100644 --- a/net/openvswitch/conntrack.c +++ b/net/openvswitch/conntrack.c @@ -514,7 +514,7 @@ static int ovs_ct_nat_execute(struct sk_buff *skb, struct nf_conn *ct, int hooknum, nh_off, err = NF_ACCEPT; nh_off = skb_network_offset(skb); - skb_pull(skb, nh_off); + skb_pull_rcsum(skb, nh_off); /* See HOOK2MANIP(). */ if (maniptype == NF_NAT_MANIP_SRC) @@ -579,6 +579,7 @@ static int ovs_ct_nat_execute(struct sk_buff *skb, struct nf_conn *ct, err = nf_nat_packet(ct, ctinfo, hooknum, skb); push: skb_push(skb, nh_off); + skb_postpush_rcsum(skb, skb->data, nh_off); return err; } @@ -890,7 +891,7 @@ int ovs_ct_execute(struct net *net, struct sk_buff *skb, /* The conntrack module expects to be working at L3. */ nh_ofs = skb_network_offset(skb); - skb_pull(skb, nh_ofs); + skb_pull_rcsum(skb, nh_ofs); if (key->ip.frag != OVS_FRAG_TYPE_NONE) { err = handle_fragments(net, key, info->zone.id, skb); @@ -904,6 +905,7 @@ int ovs_ct_execute(struct net *net, struct sk_buff *skb, err = ovs_ct_lookup(net, key, info, skb); skb_push(skb, nh_ofs); + skb_postpush_rcsum(skb, skb->data, nh_ofs); if (err) kfree_skb(skb); return err; From 77ce30dc4df48696a5b3bf8512cc2d38918116c5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 13 Jan 2017 08:39:24 -0800 Subject: [PATCH 078/119] mlx4: do not call napi_schedule() without care [ Upstream commit 8cf699ec849f4ca1413cea01289bd7d37dbcc626 ] Disable BH around the call to napi_schedule() to avoid following warning [ 52.095499] NOHZ: local_softirq_pending 08 [ 52.421291] NOHZ: local_softirq_pending 08 [ 52.608313] NOHZ: local_softirq_pending 08 Fixes: 8d59de8f7bb3 ("net/mlx4_en: Process all completions in RX rings after port goes up") Signed-off-by: Eric Dumazet Cc: Erez Shitrit Cc: Eugenia Emantayev Cc: Tariq Toukan Acked-by: Tariq Toukan Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c index fb8bb027b69c..d223e7cb68ba 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -1740,8 +1740,11 @@ int mlx4_en_start_port(struct net_device *dev) /* Process all completions if exist to prevent * the queues freezing if they are full */ - for (i = 0; i < priv->rx_ring_num; i++) + for (i = 0; i < priv->rx_ring_num; i++) { + local_bh_disable(); napi_schedule(&priv->rx_cq[i]->napi); + local_bh_enable(); + } netif_tx_start_all_queues(dev); netif_device_attach(dev); From 08e650708dd094213d1601617dd71fb8b965802f Mon Sep 17 00:00:00 2001 From: Masaru Nagai Date: Mon, 16 Jan 2017 11:45:21 +0100 Subject: [PATCH 079/119] ravb: do not use zero-length alignment DMA descriptor [ Upstream commit 8ec3e8a192ba6f13be4522ee81227c792c86fb1a ] Due to alignment requirements of the hardware transmissions are split into two DMA descriptors, a small padding descriptor of 0 - 3 bytes in length followed by a descriptor for rest of the packet. In the case of IP packets the first descriptor will never be zero due to the way that the stack aligns buffers for IP packets. However, for non-IP packets it may be zero. In that case it has been reported that timeouts occur, presumably because transmission stops at the first zero-length DMA descriptor and thus the packet is not transmitted. However, in my environment a BUG is triggered as follows: [ 20.381417] ------------[ cut here ]------------ [ 20.386054] kernel BUG at lib/swiotlb.c:495! [ 20.390324] Internal error: Oops - BUG: 0 [#1] PREEMPT SMP [ 20.395805] Modules linked in: [ 20.398862] CPU: 0 PID: 2089 Comm: mz Not tainted 4.10.0-rc3-00001-gf13ad2db193f #162 [ 20.406689] Hardware name: Renesas Salvator-X board based on r8a7796 (DT) [ 20.413474] task: ffff80063b1f1900 task.stack: ffff80063a71c000 [ 20.419404] PC is at swiotlb_tbl_map_single+0x178/0x2ec [ 20.424625] LR is at map_single+0x4c/0x98 [ 20.428629] pc : [] lr : [] pstate: 800001c5 [ 20.436019] sp : ffff80063a71f9b0 [ 20.439327] x29: ffff80063a71f9b0 x28: ffff80063a20d500 [ 20.444636] x27: ffff000008ed5000 x26: 0000000000000000 [ 20.449944] x25: 000000067abe2adc x24: 0000000000000000 [ 20.455252] x23: 0000000000200000 x22: 0000000000000001 [ 20.460559] x21: 0000000000175ffe x20: ffff80063b2a0010 [ 20.465866] x19: 0000000000000000 x18: 0000ffffcae6fb20 [ 20.471173] x17: 0000ffffa09ba018 x16: ffff0000087c8b70 [ 20.476480] x15: 0000ffffa084f588 x14: 0000ffffa09cfa14 [ 20.481787] x13: 0000ffffcae87ff0 x12: 000000000063abe2 [ 20.487098] x11: ffff000008096360 x10: ffff80063abe2adc [ 20.492407] x9 : 0000000000000000 x8 : 0000000000000000 [ 20.497718] x7 : 0000000000000000 x6 : ffff000008ed50d0 [ 20.503028] x5 : 0000000000000000 x4 : 0000000000000001 [ 20.508338] x3 : 0000000000000000 x2 : 000000067abe2adc [ 20.513648] x1 : 00000000bafff000 x0 : 0000000000000000 [ 20.518958] [ 20.520446] Process mz (pid: 2089, stack limit = 0xffff80063a71c000) [ 20.526798] Stack: (0xffff80063a71f9b0 to 0xffff80063a720000) [ 20.532543] f9a0: ffff80063a71fa30 ffff00000839c680 [ 20.540374] f9c0: ffff80063b2a0010 ffff80063b2a0010 0000000000000001 0000000000000000 [ 20.548204] f9e0: 000000000000006e ffff80063b23c000 ffff80063b23c000 0000000000000000 [ 20.556034] fa00: ffff80063b23c000 ffff80063a20d500 000000013b1f1900 0000000000000000 [ 20.563864] fa20: ffff80063ffd18e0 ffff80063b2a0010 ffff80063a71fa60 ffff00000839cd10 [ 20.571694] fa40: ffff80063b2a0010 0000000000000000 ffff80063ffd18e0 000000067abe2adc [ 20.579524] fa60: ffff80063a71fa90 ffff000008096380 ffff80063b2a0010 0000000000000000 [ 20.587353] fa80: 0000000000000000 0000000000000001 ffff80063a71fac0 ffff00000864f770 [ 20.595184] faa0: ffff80063b23caf0 0000000000000000 0000000000000000 0000000000000140 [ 20.603014] fac0: ffff80063a71fb60 ffff0000087e6498 ffff80063a20d500 ffff80063b23c000 [ 20.610843] fae0: 0000000000000000 ffff000008daeaf0 0000000000000000 ffff000008daeb00 [ 20.618673] fb00: ffff80063a71fc0c ffff000008da7000 ffff80063b23c090 ffff80063a44f000 [ 20.626503] fb20: 0000000000000000 ffff000008daeb00 ffff80063a71fc0c ffff000008da7000 [ 20.634333] fb40: ffff80063b23c090 0000000000000000 ffff800600000037 ffff0000087e63d8 [ 20.642163] fb60: ffff80063a71fbc0 ffff000008807510 ffff80063a692400 ffff80063a20d500 [ 20.649993] fb80: ffff80063a44f000 ffff80063b23c000 ffff80063a69249c 0000000000000000 [ 20.657823] fba0: 0000000000000000 ffff80063a087800 ffff80063b23c000 ffff80063a20d500 [ 20.665653] fbc0: ffff80063a71fc10 ffff0000087e67dc ffff80063a20d500 ffff80063a692400 [ 20.673483] fbe0: ffff80063b23c000 0000000000000000 ffff80063a44f000 ffff80063a69249c [ 20.681312] fc00: ffff80063a5f1a10 000000103a087800 ffff80063a71fc70 ffff0000087e6b24 [ 20.689142] fc20: ffff80063a5f1a80 ffff80063a71fde8 000000000000000f 00000000000005ea [ 20.696972] fc40: ffff80063a5f1a10 0000000000000000 000000000000000f ffff00000887fbd0 [ 20.704802] fc60: fffffff43a5f1a80 0000000000000000 ffff80063a71fc80 ffff000008880240 [ 20.712632] fc80: ffff80063a71fd90 ffff0000087c7a34 ffff80063afc7180 0000000000000000 [ 20.720462] fca0: 0000ffffcae6fe18 0000000000000014 0000000060000000 0000000000000015 [ 20.728292] fcc0: 0000000000000123 00000000000000ce ffff0000088d2000 ffff80063b1f1900 [ 20.736122] fce0: 0000000000008933 ffff000008e7cb80 ffff80063a71fd80 ffff0000087c50a4 [ 20.743951] fd00: 0000000000008933 ffff000008e7cb80 ffff000008e7cb80 000000100000000e [ 20.751781] fd20: ffff80063a71fe4c 0000ffff00000300 0000000000000123 0000000000000000 [ 20.759611] fd40: 0000000000000000 ffff80063b1f0000 000000000000000e 0000000000000300 [ 20.767441] fd60: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 [ 20.775271] fd80: 0000000000000000 0000000000000000 ffff80063a71fda0 ffff0000087c8c20 [ 20.783100] fda0: 0000000000000000 ffff000008082f30 0000000000000000 0000800637260000 [ 20.790930] fdc0: ffffffffffffffff 0000ffffa0903078 0000000000000000 000000001ea87232 [ 20.798760] fde0: 000000000000000f ffff80063a71fe40 ffff800600000014 ffff000000000001 [ 20.806590] fe00: 0000000000000000 0000000000000000 ffff80063a71fde8 0000000000000000 [ 20.814420] fe20: 0000000000000000 0000000000000000 0000000000000000 0000000000000001 [ 20.822249] fe40: 0000000203000011 0000000000000000 0000000000000000 ffff80063a68aa00 [ 20.830079] fe60: ffff80063a68aa00 0000000000000003 0000000000008933 ffff0000081f1b9c [ 20.837909] fe80: 0000000000000000 ffff000008082f30 0000000000000000 0000800637260000 [ 20.845739] fea0: ffffffffffffffff 0000ffffa07ca81c 0000000060000000 0000000000000015 [ 20.853569] fec0: 0000000000000003 000000001ea87232 000000000000000f 0000000000000000 [ 20.861399] fee0: 0000ffffcae6fe18 0000000000000014 0000000000000300 0000000000000000 [ 20.869228] ff00: 00000000000000ce 0000000000000000 00000000ffffffff 0000000000000000 [ 20.877059] ff20: 0000000000000002 0000ffffcae87ff0 0000ffffa09cfa14 0000ffffa084f588 [ 20.884888] ff40: 0000000000000000 0000ffffa09ba018 0000ffffcae6fb20 000000001ea87010 [ 20.892718] ff60: 0000ffffa09b9000 0000ffffcae6fe30 0000ffffcae6fe18 000000000000000f [ 20.900548] ff80: 0000000000000003 000000001ea87232 0000000000000000 0000000000000000 [ 20.908378] ffa0: 0000000000000000 0000ffffcae6fdc0 0000ffffa09a7824 0000ffffcae6fdc0 [ 20.916208] ffc0: 0000ffffa0903078 0000000060000000 0000000000000003 00000000000000ce [ 20.924038] ffe0: 0000000000000000 0000000000000000 ffffffffffffffff ffffffffffffffff [ 20.931867] Call trace: [ 20.934312] Exception stack(0xffff80063a71f7e0 to 0xffff80063a71f910) [ 20.940750] f7e0: 0000000000000000 0001000000000000 ffff80063a71f9b0 ffff00000839c4c0 [ 20.948580] f800: ffff80063a71f840 ffff00000888a6e4 ffff80063a24c418 ffff80063a24c448 [ 20.956410] f820: 0000000000000000 ffff00000811cd54 ffff80063a71f860 ffff80063a24c458 [ 20.964240] f840: ffff80063a71f870 ffff00000888b258 ffff80063a24c418 0000000000000001 [ 20.972070] f860: ffff80063a71f910 ffff80063a7b7028 ffff80063a71f890 ffff0000088825e4 [ 20.979899] f880: 0000000000000000 00000000bafff000 000000067abe2adc 0000000000000000 [ 20.987729] f8a0: 0000000000000001 0000000000000000 ffff000008ed50d0 0000000000000000 [ 20.995560] f8c0: 0000000000000000 0000000000000000 ffff80063abe2adc ffff000008096360 [ 21.003390] f8e0: 000000000063abe2 0000ffffcae87ff0 0000ffffa09cfa14 0000ffffa084f588 [ 21.011219] f900: ffff0000087c8b70 0000ffffa09ba018 [ 21.016097] [] swiotlb_tbl_map_single+0x178/0x2ec [ 21.022362] [] map_single+0x4c/0x98 [ 21.027411] [] swiotlb_map_page+0xa4/0x138 [ 21.033072] [] __swiotlb_map_page+0x20/0x7c [ 21.038821] [] ravb_start_xmit+0x174/0x668 [ 21.044484] [] dev_hard_start_xmit+0x8c/0x120 [ 21.050407] [] sch_direct_xmit+0x108/0x1a0 [ 21.056064] [] __dev_queue_xmit+0x194/0x4cc [ 21.061807] [] dev_queue_xmit+0x10/0x18 [ 21.067214] [] packet_sendmsg+0xf40/0x1220 [ 21.072873] [] sock_sendmsg+0x18/0x2c [ 21.078097] [] SyS_sendto+0xb0/0xf0 [ 21.083150] [] el0_svc_naked+0x24/0x28 [ 21.088462] Code: d34bfef7 2a1803f3 1a9f86d6 35fff878 (d4210000) [ 21.094611] ---[ end trace 5bc544ad491f3814 ]--- [ 21.099234] Kernel panic - not syncing: Fatal exception in interrupt [ 21.105587] Kernel Offset: disabled [ 21.109073] Memory Limit: none [ 21.112126] ---[ end Kernel panic - not syncing: Fatal exception in interrupt Fixes: 2f45d1902acf ("ravb: minimize TX data copying") Signed-off-by: Masaru Nagai Acked-by: Sergei Shtylyov Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/renesas/ravb_main.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index d6a217874a8b..862f18ed6022 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -1508,6 +1508,19 @@ static netdev_tx_t ravb_start_xmit(struct sk_buff *skb, struct net_device *ndev) buffer = PTR_ALIGN(priv->tx_align[q], DPTR_ALIGN) + entry / NUM_TX_DESC * DPTR_ALIGN; len = PTR_ALIGN(skb->data, DPTR_ALIGN) - skb->data; + /* Zero length DMA descriptors are problematic as they seem to + * terminate DMA transfers. Avoid them by simply using a length of + * DPTR_ALIGN (4) when skb data is aligned to DPTR_ALIGN. + * + * As skb is guaranteed to have at least ETH_ZLEN (60) bytes of + * data by the call to skb_put_padto() above this is safe with + * respect to both the length of the first DMA descriptor (len) + * overflowing the available data and the length of the second DMA + * descriptor (skb->len - len) being negative. + */ + if (len == 0) + len = DPTR_ALIGN; + memcpy(buffer, skb->data, len); dma_addr = dma_map_single(ndev->dev.parent, buffer, len, DMA_TO_DEVICE); if (dma_mapping_error(ndev->dev.parent, dma_addr)) From c7a5df92d1e87884a170c0245e92ac6c8f9ec4d7 Mon Sep 17 00:00:00 2001 From: Jakub Sitnicki Date: Fri, 13 Jan 2017 10:12:20 +0100 Subject: [PATCH 080/119] ip6_tunnel: Account for tunnel header in tunnel MTU [ Upstream commit 02ca0423fd65a0a9c4d70da0dbb8f4b8503f08c7 ] With ip6gre we have a tunnel header which also makes the tunnel MTU smaller. We need to reserve room for it. Previously we were using up space reserved for the Tunnel Encapsulation Limit option header (RFC 2473). Also, after commit b05229f44228 ("gre6: Cleanup GREv6 transmit path, call common GRE functions") our contract with the caller has changed. Now we check if the packet length exceeds the tunnel MTU after the tunnel header has been pushed, unlike before. This is reflected in the check where we look at the packet length minus the size of the tunnel header, which is already accounted for in tunnel MTU. Fixes: b05229f44228 ("gre6: Cleanup GREv6 transmit path, call common GRE functions") Signed-off-by: Jakub Sitnicki Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/ipv6/ip6_tunnel.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index d76674efe523..f95437f1087c 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -1108,7 +1108,7 @@ route_lookup: t->parms.name); goto tx_err_dst_release; } - mtu = dst_mtu(dst) - psh_hlen; + mtu = dst_mtu(dst) - psh_hlen - t->tun_hlen; if (encap_limit >= 0) { max_headroom += 8; mtu -= 8; @@ -1117,7 +1117,7 @@ route_lookup: mtu = IPV6_MIN_MTU; if (skb_dst(skb) && !t->parms.collect_md) skb_dst(skb)->ops->update_pmtu(skb_dst(skb), NULL, skb, mtu); - if (skb->len > mtu && !skb_is_gso(skb)) { + if (skb->len - t->tun_hlen > mtu && !skb_is_gso(skb)) { *pmtu = mtu; err = -EMSGSIZE; goto tx_err_dst_release; From 2d6b61ec9d3aed5091d6df4703a079d109bef3c1 Mon Sep 17 00:00:00 2001 From: Basil Gunn Date: Sat, 14 Jan 2017 12:18:55 -0800 Subject: [PATCH 081/119] ax25: Fix segfault after sock connection timeout [ Upstream commit 8a367e74c0120ef68c8c70d5a025648c96626dff ] The ax.25 socket connection timed out & the sock struct has been previously taken down ie. sock struct is now a NULL pointer. Checking the sock_flag causes the segfault. Check if the socket struct pointer is NULL before checking sock_flag. This segfault is seen in timed out netrom connections. Please submit to -stable. Signed-off-by: Basil Gunn Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/ax25/ax25_subr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ax25/ax25_subr.c b/net/ax25/ax25_subr.c index 655a7d4c96e1..983f0b5e14f1 100644 --- a/net/ax25/ax25_subr.c +++ b/net/ax25/ax25_subr.c @@ -264,7 +264,7 @@ void ax25_disconnect(ax25_cb *ax25, int reason) { ax25_clear_queues(ax25); - if (!sock_flag(ax25->sk, SOCK_DESTROY)) + if (!ax25->sk || !sock_flag(ax25->sk, SOCK_DESTROY)) ax25_stop_heartbeat(ax25); ax25_stop_t1timer(ax25); ax25_stop_t2timer(ax25); From b260a714a638cabb72f3d386ecb66ec04e06fcce Mon Sep 17 00:00:00 2001 From: Jamal Hadi Salim Date: Sun, 15 Jan 2017 10:14:06 -0500 Subject: [PATCH 082/119] net sched actions: fix refcnt when GETing of action after bind [ Upstream commit 0faa9cb5b3836a979864a6357e01d2046884ad52 ] Demonstrating the issue: .. add a drop action $sudo $TC actions add action drop index 10 .. retrieve it $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 2 bind 0 installed 29 sec used 29 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 ... bug 1 above: reference is two. Reference is actually 1 but we forget to subtract 1. ... do a GET again and we see the same issue try a few times and nothing changes ~$ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 2 bind 0 installed 31 sec used 31 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 ... lets try to bind the action to a filter.. $ sudo $TC qdisc add dev lo ingress $ sudo $TC filter add dev lo parent ffff: protocol ip prio 1 \ u32 match ip dst 127.0.0.1/32 flowid 1:1 action gact index 10 ... and now a few GETs: $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 3 bind 1 installed 204 sec used 204 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 4 bind 1 installed 206 sec used 206 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 5 bind 1 installed 235 sec used 235 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 .... as can be observed the reference count keeps going up. After the fix $ sudo $TC actions add action drop index 10 $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 1 bind 0 installed 4 sec used 4 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 1 bind 0 installed 6 sec used 6 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 $ sudo $TC qdisc add dev lo ingress $ sudo $TC filter add dev lo parent ffff: protocol ip prio 1 \ u32 match ip dst 127.0.0.1/32 flowid 1:1 action gact index 10 $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 2 bind 1 installed 32 sec used 32 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 $ sudo $TC -s actions get action gact index 10 action order 1: gact action drop random type none pass val 0 index 10 ref 2 bind 1 installed 33 sec used 33 sec Action statistics: Sent 0 bytes 0 pkt (dropped 0, overlimits 0 requeues 0) backlog 0b 0p requeues 0 Fixes: aecc5cefc389 ("net sched actions: fix GETing actions") Signed-off-by: Jamal Hadi Salim Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/sched/act_api.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/net/sched/act_api.c b/net/sched/act_api.c index f893d180da1c..c6c2a93cc2a2 100644 --- a/net/sched/act_api.c +++ b/net/sched/act_api.c @@ -903,8 +903,6 @@ tca_action_gd(struct net *net, struct nlattr *nla, struct nlmsghdr *n, goto err; } act->order = i; - if (event == RTM_GETACTION) - act->tcfa_refcnt++; list_add_tail(&act->list, &actions); } @@ -917,7 +915,8 @@ tca_action_gd(struct net *net, struct nlattr *nla, struct nlmsghdr *n, return ret; } err: - tcf_action_destroy(&actions, 0); + if (event != RTM_GETACTION) + tcf_action_destroy(&actions, 0); return ret; } From 3eab5dd0eb19b6ebed8ef7e7477d9f3048cc78fa Mon Sep 17 00:00:00 2001 From: Rolf Neugebauer Date: Tue, 17 Jan 2017 18:13:51 +0000 Subject: [PATCH 083/119] virtio: don't set VIRTIO_NET_HDR_F_DATA_VALID on xmit [ Upstream commit 501db511397fd6efff3aa5b4e8de415b55559550 ] This patch part reverts fd2a0437dc33 and e858fae2b0b8 which introduced a subtle change in how the virtio_net flags are derived from the SKBs ip_summed field. With the above commits, the flags are set to VIRTIO_NET_HDR_F_DATA_VALID when ip_summed == CHECKSUM_UNNECESSARY, thus treating it differently to ip_summed == CHECKSUM_NONE, which should be the same. Further, the virtio spec 1.0 / CS04 explicitly says that VIRTIO_NET_HDR_F_DATA_VALID must not be set by the driver. Fixes: fd2a0437dc33 ("virtio_net: introduce virtio_net_hdr_{from,to}_skb") Fixes: e858fae2b0b8 (" virtio_net: use common code for virtio_net_hdr and skb GSO conversion") Signed-off-by: Rolf Neugebauer Acked-by: Michael S. Tsirkin Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- include/linux/virtio_net.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/include/linux/virtio_net.h b/include/linux/virtio_net.h index 1c912f85e041..40914bb396e7 100644 --- a/include/linux/virtio_net.h +++ b/include/linux/virtio_net.h @@ -91,8 +91,6 @@ static inline int virtio_net_hdr_from_skb(const struct sk_buff *skb, skb_checksum_start_offset(skb)); hdr->csum_offset = __cpu_to_virtio16(little_endian, skb->csum_offset); - } else if (skb->ip_summed == CHECKSUM_UNNECESSARY) { - hdr->flags = VIRTIO_NET_HDR_F_DATA_VALID; } /* else everything is zero */ return 0; From 1e7cbb413f63d8fa790c8dabc208ce2a02339c26 Mon Sep 17 00:00:00 2001 From: Jason Wang Date: Fri, 20 Jan 2017 14:32:42 +0800 Subject: [PATCH 084/119] virtio-net: restore VIRTIO_HDR_F_DATA_VALID on receiving [ Upstream commit 6391a4481ba0796805d6581e42f9f0418c099e34 ] Commit 501db511397f ("virtio: don't set VIRTIO_NET_HDR_F_DATA_VALID on xmit") in fact disables VIRTIO_HDR_F_DATA_VALID on receiving path too, fixing this by adding a hint (has_data_valid) and set it only on the receiving path. Cc: Rolf Neugebauer Signed-off-by: Jason Wang Acked-by: Rolf Neugebauer Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/macvtap.c | 2 +- drivers/net/tun.c | 2 +- drivers/net/virtio_net.c | 2 +- include/linux/virtio_net.h | 6 +++++- net/packet/af_packet.c | 2 +- 5 files changed, 9 insertions(+), 5 deletions(-) diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c index 7869b0651576..6f38daf2d978 100644 --- a/drivers/net/macvtap.c +++ b/drivers/net/macvtap.c @@ -827,7 +827,7 @@ static ssize_t macvtap_put_user(struct macvtap_queue *q, return -EINVAL; ret = virtio_net_hdr_from_skb(skb, &vnet_hdr, - macvtap_is_little_endian(q)); + macvtap_is_little_endian(q), true); if (ret) BUG(); diff --git a/drivers/net/tun.c b/drivers/net/tun.c index db6acecabeaa..18402d79539e 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -1374,7 +1374,7 @@ static ssize_t tun_put_user(struct tun_struct *tun, return -EINVAL; ret = virtio_net_hdr_from_skb(skb, &gso, - tun_is_little_endian(tun)); + tun_is_little_endian(tun), true); if (ret) { struct skb_shared_info *sinfo = skb_shinfo(skb); pr_err("unexpected GSO type: " diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index cbf1c613c67a..51fc0c33a62f 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -840,7 +840,7 @@ static int xmit_skb(struct send_queue *sq, struct sk_buff *skb) hdr = skb_vnet_hdr(skb); if (virtio_net_hdr_from_skb(skb, &hdr->hdr, - virtio_is_little_endian(vi->vdev))) + virtio_is_little_endian(vi->vdev), false)) BUG(); if (vi->mergeable_rx_bufs) diff --git a/include/linux/virtio_net.h b/include/linux/virtio_net.h index 40914bb396e7..f211c348e592 100644 --- a/include/linux/virtio_net.h +++ b/include/linux/virtio_net.h @@ -56,7 +56,8 @@ static inline int virtio_net_hdr_to_skb(struct sk_buff *skb, static inline int virtio_net_hdr_from_skb(const struct sk_buff *skb, struct virtio_net_hdr *hdr, - bool little_endian) + bool little_endian, + bool has_data_valid) { memset(hdr, 0, sizeof(*hdr)); @@ -91,6 +92,9 @@ static inline int virtio_net_hdr_from_skb(const struct sk_buff *skb, skb_checksum_start_offset(skb)); hdr->csum_offset = __cpu_to_virtio16(little_endian, skb->csum_offset); + } else if (has_data_valid && + skb->ip_summed == CHECKSUM_UNNECESSARY) { + hdr->flags = VIRTIO_NET_HDR_F_DATA_VALID; } /* else everything is zero */ return 0; diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index dd2332390c45..94e4a5941d89 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -1972,7 +1972,7 @@ static int __packet_rcv_vnet(const struct sk_buff *skb, { *vnet_hdr = (const struct virtio_net_hdr) { 0 }; - if (virtio_net_hdr_from_skb(skb, vnet_hdr, vio_le())) + if (virtio_net_hdr_from_skb(skb, vnet_hdr, vio_le(), true)) BUG(); return 0; From d1c95f9ce102e77875ee8febc73d2a217dd020a0 Mon Sep 17 00:00:00 2001 From: Lance Richardson Date: Mon, 16 Jan 2017 18:37:58 -0500 Subject: [PATCH 085/119] vxlan: fix byte order of vxlan-gpe port number [ Upstream commit d5ff72d9af73bc3cbaa3edb541333a851f8c7295 ] vxlan->cfg.dst_port is in network byte order, so an htons() is needed here. Also reduced comment length to stay closer to 80 column width (still slightly over, however). Fixes: e1e5314de08b ("vxlan: implement GPE") Signed-off-by: Lance Richardson Acked-by: Jiri Benc Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/vxlan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c index 2ba01ca02c9c..0fafaa9d903b 100644 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@ -2887,7 +2887,7 @@ static int vxlan_dev_configure(struct net *src_net, struct net_device *dev, memcpy(&vxlan->cfg, conf, sizeof(*conf)); if (!vxlan->cfg.dst_port) { if (conf->flags & VXLAN_F_GPE) - vxlan->cfg.dst_port = 4790; /* IANA assigned VXLAN-GPE port */ + vxlan->cfg.dst_port = htons(4790); /* IANA VXLAN-GPE port */ else vxlan->cfg.dst_port = default_port; } From 948e137ad9ba0d4ff8ebc5ee994dc26fe3ebc4f6 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 18 Jan 2017 12:12:17 -0800 Subject: [PATCH 086/119] net: fix harmonize_features() vs NETIF_F_HIGHDMA [ Upstream commit 7be2c82cfd5d28d7adb66821a992604eb6dd112e ] Ashizuka reported a highmem oddity and sent a patch for freescale fec driver. But the problem root cause is that core networking stack must ensure no skb with highmem fragment is ever sent through a device that does not assert NETIF_F_HIGHDMA in its features. We need to call illegal_highdma() from harmonize_features() regardless of CSUM checks. Fixes: ec5f06156423 ("net: Kill link between CSUM and SG features.") Signed-off-by: Eric Dumazet Cc: Pravin Shelar Reported-by: "Ashizuka, Yuusuke" Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/core/dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/core/dev.c b/net/core/dev.c index e1d731fdc72c..df51c50927ab 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2815,9 +2815,9 @@ static netdev_features_t harmonize_features(struct sk_buff *skb, if (skb->ip_summed != CHECKSUM_NONE && !can_checksum_protocol(features, type)) { features &= ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK); - } else if (illegal_highdma(skb->dev, skb)) { - features &= ~NETIF_F_SG; } + if (illegal_highdma(skb->dev, skb)) + features &= ~NETIF_F_SG; return features; } From b335e65666610c456a5fed734cdd04ec84639a52 Mon Sep 17 00:00:00 2001 From: Daniel Gonzalez Cabanelas Date: Tue, 17 Jan 2017 16:26:55 -0800 Subject: [PATCH 087/119] net: phy: bcm63xx: Utilize correct config_intr function [ Upstream commit cd33b3e0da43522ff8e8f2b2b71d3d08298512b0 ] Commit a1cba5613edf ("net: phy: Add Broadcom phy library for common interfaces") make the BCM63xx PHY driver utilize bcm_phy_config_intr() which would appear to do the right thing, except that it does not write to the MII_BCM63XX_IR register but to MII_BCM54XX_ECR which is different. This would be causing invalid link parameters and events from being generated by the PHY interrupt. Fixes: a1cba5613edf ("net: phy: Add Broadcom phy library for common interfaces") Signed-off-by: Daniel Gonzalez Cabanelas Signed-off-by: Florian Fainelli Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/phy/bcm63xx.c | 21 +++++++++++++++++++-- 1 file changed, 19 insertions(+), 2 deletions(-) diff --git a/drivers/net/phy/bcm63xx.c b/drivers/net/phy/bcm63xx.c index e741bf614c4e..b0492ef2cdaa 100644 --- a/drivers/net/phy/bcm63xx.c +++ b/drivers/net/phy/bcm63xx.c @@ -21,6 +21,23 @@ MODULE_DESCRIPTION("Broadcom 63xx internal PHY driver"); MODULE_AUTHOR("Maxime Bizon "); MODULE_LICENSE("GPL"); +static int bcm63xx_config_intr(struct phy_device *phydev) +{ + int reg, err; + + reg = phy_read(phydev, MII_BCM63XX_IR); + if (reg < 0) + return reg; + + if (phydev->interrupts == PHY_INTERRUPT_ENABLED) + reg &= ~MII_BCM63XX_IR_GMASK; + else + reg |= MII_BCM63XX_IR_GMASK; + + err = phy_write(phydev, MII_BCM63XX_IR, reg); + return err; +} + static int bcm63xx_config_init(struct phy_device *phydev) { int reg, err; @@ -55,7 +72,7 @@ static struct phy_driver bcm63xx_driver[] = { .config_aneg = genphy_config_aneg, .read_status = genphy_read_status, .ack_interrupt = bcm_phy_ack_intr, - .config_intr = bcm_phy_config_intr, + .config_intr = bcm63xx_config_intr, }, { /* same phy as above, with just a different OUI */ .phy_id = 0x002bdc00, @@ -67,7 +84,7 @@ static struct phy_driver bcm63xx_driver[] = { .config_aneg = genphy_config_aneg, .read_status = genphy_read_status, .ack_interrupt = bcm_phy_ack_intr, - .config_intr = bcm_phy_config_intr, + .config_intr = bcm63xx_config_intr, } }; module_phy_driver(bcm63xx_driver); From e9db042dca20f7b6c1969fa21b121dde01057a74 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Tue, 17 Jan 2017 14:57:36 -0800 Subject: [PATCH 088/119] lwtunnel: fix autoload of lwt modules [ Upstream commit 9ed59592e3e379b2e9557dc1d9e9ec8fcbb33f16] Trying to add an mpls encap route when the MPLS modules are not loaded hangs. For example: CONFIG_MPLS=y CONFIG_NET_MPLS_GSO=m CONFIG_MPLS_ROUTING=m CONFIG_MPLS_IPTUNNEL=m $ ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 The ip command hangs: root 880 826 0 21:25 pts/0 00:00:00 ip route add 10.10.10.10/32 encap mpls 100 via inet 10.100.1.2 $ cat /proc/880/stack [] call_usermodehelper_exec+0xd6/0x134 [] __request_module+0x27b/0x30a [] lwtunnel_build_state+0xe4/0x178 [] fib_create_info+0x47f/0xdd4 [] fib_table_insert+0x90/0x41f [] inet_rtm_newroute+0x4b/0x52 ... modprobe is trying to load rtnl-lwt-MPLS: root 881 5 0 21:25 ? 00:00:00 /sbin/modprobe -q -- rtnl-lwt-MPLS and it hangs after loading mpls_router: $ cat /proc/881/stack [] rtnl_lock+0x12/0x14 [] register_netdevice_notifier+0x16/0x179 [] mpls_init+0x25/0x1000 [mpls_router] [] do_one_initcall+0x8e/0x13f [] do_init_module+0x5a/0x1e5 [] load_module+0x13bd/0x17d6 ... The problem is that lwtunnel_build_state is called with rtnl lock held preventing mpls_init from registering. Given the potential references held by the time lwtunnel_build_state it can not drop the rtnl lock to the load module. So, extract the module loading code from lwtunnel_build_state into a new function to validate the encap type. The new function is called while converting the user request into a fib_config which is well before any table, device or fib entries are examined. Fixes: 745041e2aaf1 ("lwtunnel: autoload of lwt modules") Signed-off-by: David Ahern Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- include/net/lwtunnel.h | 11 ++++++ net/core/lwtunnel.c | 74 ++++++++++++++++++++++++++++++++++------- net/ipv4/fib_frontend.c | 8 +++++ net/ipv6/route.c | 12 ++++++- 4 files changed, 92 insertions(+), 13 deletions(-) diff --git a/include/net/lwtunnel.h b/include/net/lwtunnel.h index ea3f80f58fd6..95581ca69f0f 100644 --- a/include/net/lwtunnel.h +++ b/include/net/lwtunnel.h @@ -106,6 +106,8 @@ int lwtunnel_encap_add_ops(const struct lwtunnel_encap_ops *op, unsigned int num); int lwtunnel_encap_del_ops(const struct lwtunnel_encap_ops *op, unsigned int num); +int lwtunnel_valid_encap_type(u16 encap_type); +int lwtunnel_valid_encap_type_attr(struct nlattr *attr, int len); int lwtunnel_build_state(struct net_device *dev, u16 encap_type, struct nlattr *encap, unsigned int family, const void *cfg, @@ -169,6 +171,15 @@ static inline int lwtunnel_encap_del_ops(const struct lwtunnel_encap_ops *op, return -EOPNOTSUPP; } +static inline int lwtunnel_valid_encap_type(u16 encap_type) +{ + return -EOPNOTSUPP; +} +static inline int lwtunnel_valid_encap_type_attr(struct nlattr *attr, int len) +{ + return -EOPNOTSUPP; +} + static inline int lwtunnel_build_state(struct net_device *dev, u16 encap_type, struct nlattr *encap, unsigned int family, const void *cfg, diff --git a/net/core/lwtunnel.c b/net/core/lwtunnel.c index e5f84c26ba1a..2f6db386624e 100644 --- a/net/core/lwtunnel.c +++ b/net/core/lwtunnel.c @@ -26,6 +26,7 @@ #include #include #include +#include #ifdef CONFIG_MODULES @@ -110,18 +111,6 @@ int lwtunnel_build_state(struct net_device *dev, u16 encap_type, ret = -EOPNOTSUPP; rcu_read_lock(); ops = rcu_dereference(lwtun_encaps[encap_type]); -#ifdef CONFIG_MODULES - if (!ops) { - const char *encap_type_str = lwtunnel_encap_str(encap_type); - - if (encap_type_str) { - rcu_read_unlock(); - request_module("rtnl-lwt-%s", encap_type_str); - rcu_read_lock(); - ops = rcu_dereference(lwtun_encaps[encap_type]); - } - } -#endif if (likely(ops && ops->build_state)) ret = ops->build_state(dev, encap, family, cfg, lws); rcu_read_unlock(); @@ -130,6 +119,67 @@ int lwtunnel_build_state(struct net_device *dev, u16 encap_type, } EXPORT_SYMBOL(lwtunnel_build_state); +int lwtunnel_valid_encap_type(u16 encap_type) +{ + const struct lwtunnel_encap_ops *ops; + int ret = -EINVAL; + + if (encap_type == LWTUNNEL_ENCAP_NONE || + encap_type > LWTUNNEL_ENCAP_MAX) + return ret; + + rcu_read_lock(); + ops = rcu_dereference(lwtun_encaps[encap_type]); + rcu_read_unlock(); +#ifdef CONFIG_MODULES + if (!ops) { + const char *encap_type_str = lwtunnel_encap_str(encap_type); + + if (encap_type_str) { + __rtnl_unlock(); + request_module("rtnl-lwt-%s", encap_type_str); + rtnl_lock(); + + rcu_read_lock(); + ops = rcu_dereference(lwtun_encaps[encap_type]); + rcu_read_unlock(); + } + } +#endif + return ops ? 0 : -EOPNOTSUPP; +} +EXPORT_SYMBOL(lwtunnel_valid_encap_type); + +int lwtunnel_valid_encap_type_attr(struct nlattr *attr, int remaining) +{ + struct rtnexthop *rtnh = (struct rtnexthop *)attr; + struct nlattr *nla_entype; + struct nlattr *attrs; + struct nlattr *nla; + u16 encap_type; + int attrlen; + + while (rtnh_ok(rtnh, remaining)) { + attrlen = rtnh_attrlen(rtnh); + if (attrlen > 0) { + attrs = rtnh_attrs(rtnh); + nla = nla_find(attrs, attrlen, RTA_ENCAP); + nla_entype = nla_find(attrs, attrlen, RTA_ENCAP_TYPE); + + if (nla_entype) { + encap_type = nla_get_u16(nla_entype); + + if (lwtunnel_valid_encap_type(encap_type) != 0) + return -EOPNOTSUPP; + } + } + rtnh = rtnh_next(rtnh, &remaining); + } + + return 0; +} +EXPORT_SYMBOL(lwtunnel_valid_encap_type_attr); + int lwtunnel_fill_encap(struct sk_buff *skb, struct lwtunnel_state *lwtstate) { const struct lwtunnel_encap_ops *ops; diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c index 3e4f183fc241..5b03d7f3b255 100644 --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -46,6 +46,7 @@ #include #include #include +#include #include #ifndef CONFIG_IP_MULTIPLE_TABLES @@ -676,6 +677,10 @@ static int rtm_to_fib_config(struct net *net, struct sk_buff *skb, cfg->fc_mx_len = nla_len(attr); break; case RTA_MULTIPATH: + err = lwtunnel_valid_encap_type_attr(nla_data(attr), + nla_len(attr)); + if (err < 0) + goto errout; cfg->fc_mp = nla_data(attr); cfg->fc_mp_len = nla_len(attr); break; @@ -690,6 +695,9 @@ static int rtm_to_fib_config(struct net *net, struct sk_buff *skb, break; case RTA_ENCAP_TYPE: cfg->fc_encap_type = nla_get_u16(attr); + err = lwtunnel_valid_encap_type(cfg->fc_encap_type); + if (err < 0) + goto errout; break; } } diff --git a/net/ipv6/route.c b/net/ipv6/route.c index acd8023763cb..bff4460f17be 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -2885,6 +2885,11 @@ static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh, if (tb[RTA_MULTIPATH]) { cfg->fc_mp = nla_data(tb[RTA_MULTIPATH]); cfg->fc_mp_len = nla_len(tb[RTA_MULTIPATH]); + + err = lwtunnel_valid_encap_type_attr(cfg->fc_mp, + cfg->fc_mp_len); + if (err < 0) + goto errout; } if (tb[RTA_PREF]) { @@ -2898,9 +2903,14 @@ static int rtm_to_fib6_config(struct sk_buff *skb, struct nlmsghdr *nlh, if (tb[RTA_ENCAP]) cfg->fc_encap = tb[RTA_ENCAP]; - if (tb[RTA_ENCAP_TYPE]) + if (tb[RTA_ENCAP_TYPE]) { cfg->fc_encap_type = nla_get_u16(tb[RTA_ENCAP_TYPE]); + err = lwtunnel_valid_encap_type(cfg->fc_encap_type); + if (err < 0) + goto errout; + } + if (tb[RTA_EXPIRES]) { unsigned long timeout = addrconf_timeout_fixup(nla_get_u32(tb[RTA_EXPIRES]), HZ); From 79453ab8816401626be91070387ec751f508219b Mon Sep 17 00:00:00 2001 From: Kefeng Wang Date: Thu, 19 Jan 2017 16:26:21 +0800 Subject: [PATCH 089/119] ipv6: addrconf: Avoid addrconf_disable_change() using RCU read-side lock [ Upstream commit 03e4deff4987f79c34112c5ba4eb195d4f9382b0 ] Just like commit 4acd4945cd1e ("ipv6: addrconf: Avoid calling netdevice notifiers with RCU read-side lock"), it is unnecessary to make addrconf_disable_change() use RCU iteration over the netdev list, since it already holds the RTNL lock, or we may meet Illegal context switch in RCU read-side critical section. Signed-off-by: Kefeng Wang Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/ipv6/addrconf.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index 4bc5ba3ae452..95dfcba38ff6 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -5515,8 +5515,7 @@ static void addrconf_disable_change(struct net *net, __s32 newf) struct net_device *dev; struct inet6_dev *idev; - rcu_read_lock(); - for_each_netdev_rcu(net, dev) { + for_each_netdev(net, dev) { idev = __in6_dev_get(dev); if (idev) { int changed = (!idev->cnf.disable_ipv6) ^ (!newf); @@ -5525,7 +5524,6 @@ static void addrconf_disable_change(struct net *net, __s32 newf) dev_disable_change(idev); } } - rcu_read_unlock(); } static int addrconf_disable_ipv6(struct ctl_table *table, int *p, int newf) From 0c687a735cdb1d7670097ab12e505a14fdec55ca Mon Sep 17 00:00:00 2001 From: Alexey Kodanev Date: Thu, 19 Jan 2017 16:36:39 +0300 Subject: [PATCH 090/119] tcp: initialize max window for a new fastopen socket [ Upstream commit 0dbd7ff3ac5017a46033a9d0a87a8267d69119d9 ] Found that if we run LTP netstress test with large MSS (65K), the first attempt from server to send data comparable to this MSS on fastopen connection will be delayed by the probe timer. Here is an example: < S seq 0:0 win 43690 options [mss 65495 wscale 7 tfo cookie] length 32 > S. seq 0:0 ack 1 win 43690 options [mss 65495 wscale 7] length 0 < . ack 1 win 342 length 0 Inside tcp_sendmsg(), tcp_send_mss() returns max MSS in 'mss_now', as well as in 'size_goal'. This results the segment not queued for transmition until all the data copied from user buffer. Then, inside __tcp_push_pending_frames(), it breaks on send window test and continues with the check probe timer. Fragmentation occurs in tcp_write_wakeup()... +0.2 > P. seq 1:43777 ack 1 win 342 length 43776 < . ack 43777, win 1365 length 0 > P. seq 43777:65001 ack 1 win 342 options [...] length 21224 ... This also contradicts with the fact that we should bound to the half of the window if it is large. Fix this flaw by correctly initializing max_window. Before that, it could have large values that affect further calculations of 'size_goal'. Fixes: 168a8f58059a ("tcp: TCP Fast Open Server - main code path") Signed-off-by: Alexey Kodanev Acked-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/ipv4/tcp_fastopen.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c index f51919535ca7..dd2560c83a85 100644 --- a/net/ipv4/tcp_fastopen.c +++ b/net/ipv4/tcp_fastopen.c @@ -205,6 +205,7 @@ static struct sock *tcp_fastopen_create_child(struct sock *sk, * scaled. So correct it appropriately. */ tp->snd_wnd = ntohs(tcp_hdr(skb)->window); + tp->max_window = tp->snd_wnd; /* Activate the retrans timer so that SYNACK can be retransmitted. * The request socket is not added to the ehash From 087dced61adcd8275de2586aa1d9f51c3887c1bf Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 18 Jan 2017 23:03:08 -0800 Subject: [PATCH 091/119] net/mlx5e: Do not recycle pages from emergency reserve [ Upstream commit e048fc50d7bde23136e098e04a324d7e3404408d ] A driver using dev_alloc_page() must not reuse a page allocated from emergency memory reserve. Otherwise all packets using this page will be immediately dropped, unless for very specific sockets having SOCK_MEMALLOC bit set. This issue might be hard to debug, because only a fraction of received packets would be dropped. Fixes: 4415a0319f92 ("net/mlx5e: Implement RX mapped page cache for page recycle") Signed-off-by: Eric Dumazet Cc: Tariq Toukan Cc: Saeed Mahameed Acked-by: Saeed Mahameed Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c index 33495d88aeb2..e7b2158bb48a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -193,6 +193,9 @@ static inline bool mlx5e_rx_cache_put(struct mlx5e_rq *rq, return false; } + if (unlikely(page_is_pfmemalloc(dma_info->page))) + return false; + cache->page_cache[cache->tail] = *dma_info; cache->tail = tail_next; return true; From 74423145d931f60cde223cc8a84a1ad3bedd772a Mon Sep 17 00:00:00 2001 From: Ivan Vecera Date: Fri, 20 Jan 2017 18:12:17 +0100 Subject: [PATCH 092/119] bridge: netlink: call br_changelink() during br_dev_newlink() [ Upstream commit b6677449dff674cf5b81429b11d5c7f358852ef9 ] Any bridge options specified during link creation (e.g. ip link add) are ignored as br_dev_newlink() does not process them. Use br_changelink() to do it. Fixes: 133235161721 ("bridge: implement rtnl_link_ops->changelink") Signed-off-by: Ivan Vecera Reviewed-by: Jiri Pirko Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/bridge/br_netlink.c | 33 +++++++++++++++++++-------------- 1 file changed, 19 insertions(+), 14 deletions(-) diff --git a/net/bridge/br_netlink.c b/net/bridge/br_netlink.c index e99037c6f7b7..04741064a173 100644 --- a/net/bridge/br_netlink.c +++ b/net/bridge/br_netlink.c @@ -781,20 +781,6 @@ static int br_validate(struct nlattr *tb[], struct nlattr *data[]) return 0; } -static int br_dev_newlink(struct net *src_net, struct net_device *dev, - struct nlattr *tb[], struct nlattr *data[]) -{ - struct net_bridge *br = netdev_priv(dev); - - if (tb[IFLA_ADDRESS]) { - spin_lock_bh(&br->lock); - br_stp_change_bridge_id(br, nla_data(tb[IFLA_ADDRESS])); - spin_unlock_bh(&br->lock); - } - - return register_netdevice(dev); -} - static int br_port_slave_changelink(struct net_device *brdev, struct net_device *dev, struct nlattr *tb[], @@ -1093,6 +1079,25 @@ static int br_changelink(struct net_device *brdev, struct nlattr *tb[], return 0; } +static int br_dev_newlink(struct net *src_net, struct net_device *dev, + struct nlattr *tb[], struct nlattr *data[]) +{ + struct net_bridge *br = netdev_priv(dev); + int err; + + if (tb[IFLA_ADDRESS]) { + spin_lock_bh(&br->lock); + br_stp_change_bridge_id(br, nla_data(tb[IFLA_ADDRESS])); + spin_unlock_bh(&br->lock); + } + + err = br_changelink(dev, tb, data); + if (err) + return err; + + return register_netdevice(dev); +} + static size_t br_get_size(const struct net_device *brdev) { return nla_total_size(sizeof(u32)) + /* IFLA_BR_FORWARD_DELAY */ From ad864d9fce0ec56cc8f6afe5c6a0e6d7f484b9eb Mon Sep 17 00:00:00 2001 From: David Ahern Date: Fri, 20 Jan 2017 12:58:34 -0800 Subject: [PATCH 093/119] net: mpls: Fix multipath selection for LSR use case [ Upstream commit 9f427a0e474a67b454420c131709600d44850486 ] MPLS multipath for LSR is broken -- always selecting the first nexthop in the one label case. For example: $ ip -f mpls ro ls 100 nexthop as to 200 via inet 172.16.2.2 dev virt12 nexthop as to 300 via inet 172.16.3.2 dev virt13 101 nexthop as to 201 via inet6 2000:2::2 dev virt12 nexthop as to 301 via inet6 2000:3::2 dev virt13 In this example incoming packets have a single MPLS labels which means BOS bit is set. The BOS bit is passed from mpls_forward down to mpls_multipath_hash which never processes the hash loop because BOS is 1. Update mpls_multipath_hash to process the entire label stack. mpls_hdr_len tracks the total mpls header length on each pass (on pass N mpls_hdr_len is N * sizeof(mpls_shim_hdr)). When the label is found with the BOS set it verifies the skb has sufficient header for ipv4 or ipv6, and find the IPv4 and IPv6 header by using the last mpls_hdr pointer and adding 1 to advance past it. With these changes I have verified the code correctly sees the label, BOS, IPv4 and IPv6 addresses in the network header and icmp/tcp/udp traffic for ipv4 and ipv6 are distributed across the nexthops. Fixes: 1c78efa8319ca ("mpls: flow-based multipath selection") Acked-by: Robert Shearman Signed-off-by: David Ahern Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/mpls/af_mpls.c | 48 ++++++++++++++++++++++++---------------------- 1 file changed, 25 insertions(+), 23 deletions(-) diff --git a/net/mpls/af_mpls.c b/net/mpls/af_mpls.c index 15fe97644ffe..5b77377e5a15 100644 --- a/net/mpls/af_mpls.c +++ b/net/mpls/af_mpls.c @@ -98,18 +98,19 @@ bool mpls_pkt_too_big(const struct sk_buff *skb, unsigned int mtu) } EXPORT_SYMBOL_GPL(mpls_pkt_too_big); -static u32 mpls_multipath_hash(struct mpls_route *rt, - struct sk_buff *skb, bool bos) +static u32 mpls_multipath_hash(struct mpls_route *rt, struct sk_buff *skb) { struct mpls_entry_decoded dec; + unsigned int mpls_hdr_len = 0; struct mpls_shim_hdr *hdr; bool eli_seen = false; int label_index; u32 hash = 0; - for (label_index = 0; label_index < MAX_MP_SELECT_LABELS && !bos; + for (label_index = 0; label_index < MAX_MP_SELECT_LABELS; label_index++) { - if (!pskb_may_pull(skb, sizeof(*hdr) * label_index)) + mpls_hdr_len += sizeof(*hdr); + if (!pskb_may_pull(skb, mpls_hdr_len)) break; /* Read and decode the current label */ @@ -134,37 +135,38 @@ static u32 mpls_multipath_hash(struct mpls_route *rt, eli_seen = true; } - bos = dec.bos; - if (bos && pskb_may_pull(skb, sizeof(*hdr) * label_index + - sizeof(struct iphdr))) { + if (!dec.bos) + continue; + + /* found bottom label; does skb have room for a header? */ + if (pskb_may_pull(skb, mpls_hdr_len + sizeof(struct iphdr))) { const struct iphdr *v4hdr; - v4hdr = (const struct iphdr *)(mpls_hdr(skb) + - label_index); + v4hdr = (const struct iphdr *)(hdr + 1); if (v4hdr->version == 4) { hash = jhash_3words(ntohl(v4hdr->saddr), ntohl(v4hdr->daddr), v4hdr->protocol, hash); } else if (v4hdr->version == 6 && - pskb_may_pull(skb, sizeof(*hdr) * label_index + - sizeof(struct ipv6hdr))) { + pskb_may_pull(skb, mpls_hdr_len + + sizeof(struct ipv6hdr))) { const struct ipv6hdr *v6hdr; - v6hdr = (const struct ipv6hdr *)(mpls_hdr(skb) + - label_index); - + v6hdr = (const struct ipv6hdr *)(hdr + 1); hash = __ipv6_addr_jhash(&v6hdr->saddr, hash); hash = __ipv6_addr_jhash(&v6hdr->daddr, hash); hash = jhash_1word(v6hdr->nexthdr, hash); } } + + break; } return hash; } static struct mpls_nh *mpls_select_multipath(struct mpls_route *rt, - struct sk_buff *skb, bool bos) + struct sk_buff *skb) { int alive = ACCESS_ONCE(rt->rt_nhn_alive); u32 hash = 0; @@ -180,7 +182,7 @@ static struct mpls_nh *mpls_select_multipath(struct mpls_route *rt, if (alive <= 0) return NULL; - hash = mpls_multipath_hash(rt, skb, bos); + hash = mpls_multipath_hash(rt, skb); nh_index = hash % alive; if (alive == rt->rt_nhn) goto out; @@ -278,17 +280,11 @@ static int mpls_forward(struct sk_buff *skb, struct net_device *dev, hdr = mpls_hdr(skb); dec = mpls_entry_decode(hdr); - /* Pop the label */ - skb_pull(skb, sizeof(*hdr)); - skb_reset_network_header(skb); - - skb_orphan(skb); - rt = mpls_route_input_rcu(net, dec.label); if (!rt) goto drop; - nh = mpls_select_multipath(rt, skb, dec.bos); + nh = mpls_select_multipath(rt, skb); if (!nh) goto drop; @@ -297,6 +293,12 @@ static int mpls_forward(struct sk_buff *skb, struct net_device *dev, if (!mpls_output_possible(out_dev)) goto drop; + /* Pop the label */ + skb_pull(skb, sizeof(*hdr)); + skb_reset_network_header(skb); + + skb_orphan(skb); + if (skb_warn_if_lro(skb)) goto drop; From 37b27b20a8a948ce5dfbd6e47272fd59e8bd0232 Mon Sep 17 00:00:00 2001 From: hayeswang Date: Mon, 23 Jan 2017 14:18:43 +0800 Subject: [PATCH 094/119] r8152: don't execute runtime suspend if the tx is not empty [ Upstream commit 6a0b76c04ec157c88ca943debf78a8ee58469f2d ] Runtime suspend shouldn't be executed if the tx queue is not empty, because the device is not idle. Signed-off-by: Hayes Wang Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/usb/r8152.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c index be4c86faba4f..90b426c5ffce 100644 --- a/drivers/net/usb/r8152.c +++ b/drivers/net/usb/r8152.c @@ -32,7 +32,7 @@ #define NETNEXT_VERSION "08" /* Information for net */ -#define NET_VERSION "6" +#define NET_VERSION "7" #define DRIVER_VERSION "v1." NETNEXT_VERSION "." NET_VERSION #define DRIVER_AUTHOR "Realtek linux nic maintainers " @@ -3572,6 +3572,8 @@ static bool delay_autosuspend(struct r8152 *tp) */ if (!sw_linking && tp->rtl_ops.in_nway(tp)) return true; + else if (!skb_queue_empty(&tp->tx_queue)) + return true; else return false; } From 93ff5e03bcba0761055491dc6bf52b1e0e33bbe6 Mon Sep 17 00:00:00 2001 From: WANG Cong Date: Mon, 23 Jan 2017 11:17:35 -0800 Subject: [PATCH 095/119] af_unix: move unix_mknod() out of bindlock [ Upstream commit 0fb44559ffd67de8517098b81f675fa0210f13f0 ] Dmitry reported a deadlock scenario: unix_bind() path: u->bindlock ==> sb_writer do_splice() path: sb_writer ==> pipe->mutex ==> u->bindlock In the unix_bind() code path, unix_mknod() does not have to be done with u->bindlock held, since it is a pure fs operation, so we can just move unix_mknod() out. Reported-by: Dmitry Vyukov Tested-by: Dmitry Vyukov Cc: Rainer Weikusat Cc: Al Viro Signed-off-by: Cong Wang Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/unix/af_unix.c | 27 ++++++++++++++++----------- 1 file changed, 16 insertions(+), 11 deletions(-) diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c index 2358f2690ec5..2d03d5bcb5b9 100644 --- a/net/unix/af_unix.c +++ b/net/unix/af_unix.c @@ -995,6 +995,7 @@ static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) unsigned int hash; struct unix_address *addr; struct hlist_head *list; + struct path path = { NULL, NULL }; err = -EINVAL; if (sunaddr->sun_family != AF_UNIX) @@ -1010,9 +1011,20 @@ static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) goto out; addr_len = err; + if (sun_path[0]) { + umode_t mode = S_IFSOCK | + (SOCK_INODE(sock)->i_mode & ~current_umask()); + err = unix_mknod(sun_path, mode, &path); + if (err) { + if (err == -EEXIST) + err = -EADDRINUSE; + goto out; + } + } + err = mutex_lock_interruptible(&u->bindlock); if (err) - goto out; + goto out_put; err = -EINVAL; if (u->addr) @@ -1029,16 +1041,6 @@ static int unix_bind(struct socket *sock, struct sockaddr *uaddr, int addr_len) atomic_set(&addr->refcnt, 1); if (sun_path[0]) { - struct path path; - umode_t mode = S_IFSOCK | - (SOCK_INODE(sock)->i_mode & ~current_umask()); - err = unix_mknod(sun_path, mode, &path); - if (err) { - if (err == -EEXIST) - err = -EADDRINUSE; - unix_release_addr(addr); - goto out_up; - } addr->hash = UNIX_HASH_SIZE; hash = d_real_inode(path.dentry)->i_ino & (UNIX_HASH_SIZE - 1); spin_lock(&unix_table_lock); @@ -1065,6 +1067,9 @@ out_unlock: spin_unlock(&unix_table_lock); out_up: mutex_unlock(&u->bindlock); +out_put: + if (err) + path_put(&path); out: return err; } From 087c2ecb21afddaeaa9105ac7df6015eb1e2ce6a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=C3=B8rn=20Mork?= Date: Tue, 24 Jan 2017 10:45:38 +0100 Subject: [PATCH 096/119] qmi_wwan/cdc_ether: add device ID for HP lt2523 (Novatel E371) WWAN card MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit [ Upstream commit 5b9f57516337b523f7466a53939aaaea7b78141b ] Another rebranded Novatel E371. qmi_wwan should drive this device, while cdc_ether should ignore it. Even though the USB descriptors are plain CDC-ETHER that USB interface is a QMI interface. Ref commit 7fdb7846c9ca ("qmi_wwan/cdc_ether: add device IDs for Dell 5804 (Novatel E371) WWAN card") Cc: Dan Williams Signed-off-by: Bjørn Mork Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- drivers/net/usb/cdc_ether.c | 8 ++++++++ drivers/net/usb/qmi_wwan.c | 7 +++++++ 2 files changed, 15 insertions(+) diff --git a/drivers/net/usb/cdc_ether.c b/drivers/net/usb/cdc_ether.c index dd623f674487..b82be816256c 100644 --- a/drivers/net/usb/cdc_ether.c +++ b/drivers/net/usb/cdc_ether.c @@ -531,6 +531,7 @@ static const struct driver_info wwan_info = { #define SAMSUNG_VENDOR_ID 0x04e8 #define LENOVO_VENDOR_ID 0x17ef #define NVIDIA_VENDOR_ID 0x0955 +#define HP_VENDOR_ID 0x03f0 static const struct usb_device_id products[] = { /* BLACKLIST !! @@ -677,6 +678,13 @@ static const struct usb_device_id products[] = { .driver_info = 0, }, +/* HP lt2523 (Novatel E371) - handled by qmi_wwan */ +{ + USB_DEVICE_AND_INTERFACE_INFO(HP_VENDOR_ID, 0x421d, USB_CLASS_COMM, + USB_CDC_SUBCLASS_ETHERNET, USB_CDC_PROTO_NONE), + .driver_info = 0, +}, + /* AnyDATA ADU960S - handled by qmi_wwan */ { USB_DEVICE_AND_INTERFACE_INFO(0x16d5, 0x650a, USB_CLASS_COMM, diff --git a/drivers/net/usb/qmi_wwan.c b/drivers/net/usb/qmi_wwan.c index 6fe1cdb0174f..24d5272cdce5 100644 --- a/drivers/net/usb/qmi_wwan.c +++ b/drivers/net/usb/qmi_wwan.c @@ -654,6 +654,13 @@ static const struct usb_device_id products[] = { USB_CDC_PROTO_NONE), .driver_info = (unsigned long)&qmi_wwan_info, }, + { /* HP lt2523 (Novatel E371) */ + USB_DEVICE_AND_INTERFACE_INFO(0x03f0, 0x421d, + USB_CLASS_COMM, + USB_CDC_SUBCLASS_ETHERNET, + USB_CDC_PROTO_NONE), + .driver_info = (unsigned long)&qmi_wwan_info, + }, { /* HP lt4112 LTE/HSPA+ Gobi 4G Module (Huawei me906e) */ USB_DEVICE_AND_INTERFACE_INFO(0x03f0, 0x581d, USB_CLASS_VENDOR_SPEC, 1, 7), .driver_info = (unsigned long)&qmi_wwan_info, From 89c2588627c20cf9d791a9bb1523646b101a59b1 Mon Sep 17 00:00:00 2001 From: Robert Shearman Date: Tue, 24 Jan 2017 16:26:47 +0000 Subject: [PATCH 097/119] net: Specify the owning module for lwtunnel ops [ Upstream commit 88ff7334f25909802140e690c0e16433e485b0a0 ] Modules implementing lwtunnel ops should not be allowed to unload while there is state alive using those ops, so specify the owning module for all lwtunnel ops. Signed-off-by: Robert Shearman Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- include/net/lwtunnel.h | 2 ++ net/ipv4/ip_tunnel_core.c | 2 ++ net/ipv6/ila/ila_lwt.c | 1 + net/mpls/mpls_iptunnel.c | 1 + 4 files changed, 6 insertions(+) diff --git a/include/net/lwtunnel.h b/include/net/lwtunnel.h index 95581ca69f0f..3e0a7a942a0e 100644 --- a/include/net/lwtunnel.h +++ b/include/net/lwtunnel.h @@ -43,6 +43,8 @@ struct lwtunnel_encap_ops { int (*get_encap_size)(struct lwtunnel_state *lwtstate); int (*cmp_encap)(struct lwtunnel_state *a, struct lwtunnel_state *b); int (*xmit)(struct sk_buff *skb); + + struct module *owner; }; #ifdef CONFIG_LWTUNNEL diff --git a/net/ipv4/ip_tunnel_core.c b/net/ipv4/ip_tunnel_core.c index fed3d29f9eb3..0fd1976ab63b 100644 --- a/net/ipv4/ip_tunnel_core.c +++ b/net/ipv4/ip_tunnel_core.c @@ -313,6 +313,7 @@ static const struct lwtunnel_encap_ops ip_tun_lwt_ops = { .fill_encap = ip_tun_fill_encap_info, .get_encap_size = ip_tun_encap_nlsize, .cmp_encap = ip_tun_cmp_encap, + .owner = THIS_MODULE, }; static const struct nla_policy ip6_tun_policy[LWTUNNEL_IP6_MAX + 1] = { @@ -403,6 +404,7 @@ static const struct lwtunnel_encap_ops ip6_tun_lwt_ops = { .fill_encap = ip6_tun_fill_encap_info, .get_encap_size = ip6_tun_encap_nlsize, .cmp_encap = ip_tun_cmp_encap, + .owner = THIS_MODULE, }; void __init ip_tunnel_core_init(void) diff --git a/net/ipv6/ila/ila_lwt.c b/net/ipv6/ila/ila_lwt.c index e50c27a93e17..f3db364fc853 100644 --- a/net/ipv6/ila/ila_lwt.c +++ b/net/ipv6/ila/ila_lwt.c @@ -164,6 +164,7 @@ static const struct lwtunnel_encap_ops ila_encap_ops = { .fill_encap = ila_fill_encap_info, .get_encap_size = ila_encap_nlsize, .cmp_encap = ila_encap_cmp, + .owner = THIS_MODULE, }; int ila_lwt_init(void) diff --git a/net/mpls/mpls_iptunnel.c b/net/mpls/mpls_iptunnel.c index cf52cf30ac4b..bc9aaf58c7cc 100644 --- a/net/mpls/mpls_iptunnel.c +++ b/net/mpls/mpls_iptunnel.c @@ -218,6 +218,7 @@ static const struct lwtunnel_encap_ops mpls_iptun_ops = { .fill_encap = mpls_fill_encap_info, .get_encap_size = mpls_encap_nlsize, .cmp_encap = mpls_encap_cmp, + .owner = THIS_MODULE, }; static int __init mpls_iptunnel_init(void) From e972cce0c833fa990622a2f46db79979ab07485c Mon Sep 17 00:00:00 2001 From: Robert Shearman Date: Tue, 24 Jan 2017 16:26:48 +0000 Subject: [PATCH 098/119] lwtunnel: Fix oops on state free after encap module unload [ Upstream commit 85c814016ce3b371016c2c054a905fa2492f5a65 ] When attempting to free lwtunnel state after the module for the encap has been unloaded an oops occurs: BUG: unable to handle kernel NULL pointer dereference at 0000000000000008 IP: lwtstate_free+0x18/0x40 [..] task: ffff88003e372380 task.stack: ffffc900001fc000 RIP: 0010:lwtstate_free+0x18/0x40 RSP: 0018:ffff88003fd83e88 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff88002bbb3380 RCX: ffff88000c91a300 [..] Call Trace: free_fib_info_rcu+0x195/0x1a0 ? rt_fibinfo_free+0x50/0x50 rcu_process_callbacks+0x2d3/0x850 ? rcu_process_callbacks+0x296/0x850 __do_softirq+0xe4/0x4cb irq_exit+0xb0/0xc0 smp_apic_timer_interrupt+0x3d/0x50 apic_timer_interrupt+0x93/0xa0 [..] Code: e8 6e c6 fc ff 89 d8 5b 5d c3 bb de ff ff ff eb f4 66 90 66 66 66 66 90 55 48 89 e5 53 0f b7 07 48 89 fb 48 8b 04 c5 00 81 d5 81 <48> 8b 40 08 48 85 c0 74 13 ff d0 48 8d 7b 20 be 20 00 00 00 e8 The problem is after the module for the encap can be unloaded the corresponding ops is removed and is thus NULL here. Modules implementing lwtunnel ops should not be allowed to unload while there is state alive using those ops, so grab the module reference for the ops on creating lwtunnel state and of course release the reference when freeing the state. Fixes: 1104d9ba443a ("lwtunnel: Add destroy state operation") Signed-off-by: Robert Shearman Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- include/net/lwtunnel.h | 5 +---- net/core/lwtunnel.c | 14 +++++++++++++- 2 files changed, 14 insertions(+), 5 deletions(-) diff --git a/include/net/lwtunnel.h b/include/net/lwtunnel.h index 3e0a7a942a0e..fc7c0dbdd1ff 100644 --- a/include/net/lwtunnel.h +++ b/include/net/lwtunnel.h @@ -48,10 +48,7 @@ struct lwtunnel_encap_ops { }; #ifdef CONFIG_LWTUNNEL -static inline void lwtstate_free(struct lwtunnel_state *lws) -{ - kfree(lws); -} +void lwtstate_free(struct lwtunnel_state *lws); static inline struct lwtunnel_state * lwtstate_get(struct lwtunnel_state *lws) diff --git a/net/core/lwtunnel.c b/net/core/lwtunnel.c index 2f6db386624e..afa64f086d87 100644 --- a/net/core/lwtunnel.c +++ b/net/core/lwtunnel.c @@ -66,6 +66,15 @@ EXPORT_SYMBOL(lwtunnel_state_alloc); static const struct lwtunnel_encap_ops __rcu * lwtun_encaps[LWTUNNEL_ENCAP_MAX + 1] __read_mostly; +void lwtstate_free(struct lwtunnel_state *lws) +{ + const struct lwtunnel_encap_ops *ops = lwtun_encaps[lws->type]; + + kfree(lws); + module_put(ops->owner); +} +EXPORT_SYMBOL(lwtstate_free); + int lwtunnel_encap_add_ops(const struct lwtunnel_encap_ops *ops, unsigned int num) { @@ -111,8 +120,11 @@ int lwtunnel_build_state(struct net_device *dev, u16 encap_type, ret = -EOPNOTSUPP; rcu_read_lock(); ops = rcu_dereference(lwtun_encaps[encap_type]); - if (likely(ops && ops->build_state)) + if (likely(ops && ops->build_state && try_module_get(ops->owner))) { ret = ops->build_state(dev, encap, family, cfg, lws); + if (ret) + module_put(ops->owner); + } rcu_read_unlock(); return ret; From 9f42bc4f9c1c968ce7f87b451544a119600e3666 Mon Sep 17 00:00:00 2001 From: Florian Fainelli Date: Wed, 25 Jan 2017 09:10:41 -0800 Subject: [PATCH 099/119] net: dsa: Bring back device detaching in dsa_slave_suspend() [ Upstream commit f154be241d22298d2b63c9b613f619fa1086ea75 ] Commit 448b4482c671 ("net: dsa: Add lockdep class to tx queues to avoid lockdep splat") removed the netif_device_detach() call done in dsa_slave_suspend() which is necessary, and paired with a corresponding netif_device_attach(), bring it back. Fixes: 448b4482c671 ("net: dsa: Add lockdep class to tx queues to avoid lockdep splat") Signed-off-by: Florian Fainelli Reviewed-by: Andrew Lunn Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- net/dsa/slave.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/dsa/slave.c b/net/dsa/slave.c index 30e2e21d7619..3ff9d97cf56b 100644 --- a/net/dsa/slave.c +++ b/net/dsa/slave.c @@ -1201,6 +1201,8 @@ int dsa_slave_suspend(struct net_device *slave_dev) { struct dsa_slave_priv *p = netdev_priv(slave_dev); + netif_device_detach(slave_dev); + if (p->phy) { phy_stop(p->phy); p->old_pause = -1; From d20e4ad06c4be123ef350a0ca78f1480bbdc5f8e Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 2 Feb 2017 08:55:53 +0100 Subject: [PATCH 100/119] xfs: bump up reserved blocks in xfs_alloc_set_aside commit 5149fd327f16e393c1d04fa5325ab072c32472bf upstream. Setting aside 4 blocks globally for bmbt splits isn't all that useful, as different threads can allocate space in parallel. Bump it to 4 blocks per AG to allow each thread that is currently doing an allocation to dip into it separately. Without that we may no have enough reserved blocks if there are enough parallel transactions in an almost out space file system that all run into bmap btree splits. Signed-off-by: Christoph Hellwig Reviewed-by: Brian Foster Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_alloc.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index 5050056a0b06..0a46f8488b8d 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -95,10 +95,7 @@ unsigned int xfs_alloc_set_aside( struct xfs_mount *mp) { - unsigned int blocks; - - blocks = 4 + (mp->m_sb.sb_agcount * XFS_ALLOC_AGFL_RESERVE); - return blocks; + return mp->m_sb.sb_agcount * (XFS_ALLOC_AGFL_RESERVE + 4); } /* From c63f4d3aa09d0b8c36836a887d32cf20a974509f Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 2 Feb 2017 08:55:54 +0100 Subject: [PATCH 101/119] xfs: fix bogus minleft manipulations commit 255c516278175a6dc7037d1406307f35237d8688 upstream. We can't just set minleft to 0 when we're low on space - that's exactly what we need minleft for: to protect space in the AG for btree block allocations when we are low on free space. Signed-off-by: Christoph Hellwig Reviewed-by: Brian Foster Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_alloc.c | 24 +++++++----------------- fs/xfs/libxfs/xfs_bmap.c | 3 --- fs/xfs/libxfs/xfs_bmap_btree.c | 3 +-- 3 files changed, 8 insertions(+), 22 deletions(-) diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index 0a46f8488b8d..fe925702c955 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -2635,12 +2635,10 @@ xfs_alloc_vextent( xfs_agblock_t agsize; /* allocation group size */ int error; int flags; /* XFS_ALLOC_FLAG_... locking flags */ - xfs_extlen_t minleft;/* minimum left value, temp copy */ xfs_mount_t *mp; /* mount structure pointer */ xfs_agnumber_t sagno; /* starting allocation group number */ xfs_alloctype_t type; /* input allocation type */ int bump_rotor = 0; - int no_min = 0; xfs_agnumber_t rotorstep = xfs_rotorstep; /* inode32 agf stepper */ mp = args->mp; @@ -2669,7 +2667,6 @@ xfs_alloc_vextent( trace_xfs_alloc_vextent_badargs(args); return 0; } - minleft = args->minleft; switch (type) { case XFS_ALLOCTYPE_THIS_AG: @@ -2680,9 +2677,7 @@ xfs_alloc_vextent( */ args->agno = XFS_FSB_TO_AGNO(mp, args->fsbno); args->pag = xfs_perag_get(mp, args->agno); - args->minleft = 0; error = xfs_alloc_fix_freelist(args, 0); - args->minleft = minleft; if (error) { trace_xfs_alloc_vextent_nofix(args); goto error0; @@ -2747,9 +2742,7 @@ xfs_alloc_vextent( */ for (;;) { args->pag = xfs_perag_get(mp, args->agno); - if (no_min) args->minleft = 0; error = xfs_alloc_fix_freelist(args, flags); - args->minleft = minleft; if (error) { trace_xfs_alloc_vextent_nofix(args); goto error0; @@ -2789,20 +2782,17 @@ xfs_alloc_vextent( * or switch to non-trylock mode. */ if (args->agno == sagno) { - if (no_min == 1) { + if (flags == 0) { args->agbno = NULLAGBLOCK; trace_xfs_alloc_vextent_allfailed(args); break; } - if (flags == 0) { - no_min = 1; - } else { - flags = 0; - if (type == XFS_ALLOCTYPE_START_BNO) { - args->agbno = XFS_FSB_TO_AGBNO(mp, - args->fsbno); - args->type = XFS_ALLOCTYPE_NEAR_BNO; - } + + flags = 0; + if (type == XFS_ALLOCTYPE_START_BNO) { + args->agbno = XFS_FSB_TO_AGBNO(mp, + args->fsbno); + args->type = XFS_ALLOCTYPE_NEAR_BNO; } } xfs_perag_put(args->pag); diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index 89d727b659fc..52dc5c175001 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -3903,7 +3903,6 @@ xfs_bmap_btalloc( args.fsbno = 0; args.type = XFS_ALLOCTYPE_FIRST_AG; args.total = ap->minlen; - args.minleft = 0; if ((error = xfs_alloc_vextent(&args))) return error; ap->dfops->dop_low = true; @@ -4437,8 +4436,6 @@ xfs_bmapi_allocate( if (error) return error; - if (bma->dfops->dop_low) - bma->minleft = 0; if (bma->cur) bma->cur->bc_private.b.firstblock = *bma->firstblock; if (bma->blkno == NULLFSBLOCK) diff --git a/fs/xfs/libxfs/xfs_bmap_btree.c b/fs/xfs/libxfs/xfs_bmap_btree.c index 049fa597ae91..f76c1693ff01 100644 --- a/fs/xfs/libxfs/xfs_bmap_btree.c +++ b/fs/xfs/libxfs/xfs_bmap_btree.c @@ -502,12 +502,11 @@ try_another_ag: if (args.fsbno == NULLFSBLOCK && args.minleft) { /* * Could not find an AG with enough free space to satisfy - * a full btree split. Try again without minleft and if + * a full btree split. Try again and if * successful activate the lowspace algorithm. */ args.fsbno = 0; args.type = XFS_ALLOCTYPE_FIRST_AG; - args.minleft = 0; error = xfs_alloc_vextent(&args); if (error) goto error0; From 6b81365b1e5795578c4aad59a37c15e958cdcdfe Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 2 Feb 2017 08:55:55 +0100 Subject: [PATCH 102/119] xfs: adjust allocation length in xfs_alloc_space_available commit 54fee133ad59c87ab01dd84ab3e9397134b32acb upstream. We must decide in xfs_alloc_fix_freelist if we can perform an allocation from a given AG is possible or not based on the available space, and should not fail the allocation past that point on a healthy file system. But currently we have two additional places that second-guess xfs_alloc_fix_freelist: xfs_alloc_ag_vextent tries to adjust the maxlen parameter to remove the reservation before doing the allocation (but ignores the various minium freespace requirements), and xfs_alloc_fix_minleft tries to fix up the allocated length after we've found an extent, but ignores the reservations and also doesn't take the AGFL into account (and thus fails allocations for not matching minlen in some cases). Remove all these later fixups and just correct the maxlen argument inside xfs_alloc_fix_freelist once we have the AGF buffer locked. Signed-off-by: Christoph Hellwig Reviewed-by: Brian Foster Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_alloc.c | 81 ++++++++------------------------------- fs/xfs/libxfs/xfs_alloc.h | 2 +- 2 files changed, 18 insertions(+), 65 deletions(-) diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index fe925702c955..f2e7eb6e5243 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -362,35 +362,11 @@ xfs_alloc_fix_len( return; ASSERT(rlen >= args->minlen && rlen <= args->maxlen); ASSERT(rlen % args->prod == args->mod); + ASSERT(args->pag->pagf_freeblks + args->pag->pagf_flcount >= + rlen + args->minleft); args->len = rlen; } -/* - * Fix up length if there is too little space left in the a.g. - * Return 1 if ok, 0 if too little, should give up. - */ -STATIC int -xfs_alloc_fix_minleft( - xfs_alloc_arg_t *args) /* allocation argument structure */ -{ - xfs_agf_t *agf; /* a.g. freelist header */ - int diff; /* free space difference */ - - if (args->minleft == 0) - return 1; - agf = XFS_BUF_TO_AGF(args->agbp); - diff = be32_to_cpu(agf->agf_freeblks) - - args->len - args->minleft; - if (diff >= 0) - return 1; - args->len += diff; /* shrink the allocated space */ - /* casts to (int) catch length underflows */ - if ((int)args->len >= (int)args->minlen) - return 1; - args->agbno = NULLAGBLOCK; - return 0; -} - /* * Update the two btrees, logically removing from freespace the extent * starting at rbno, rlen blocks. The extent is contained within the @@ -686,8 +662,6 @@ xfs_alloc_ag_vextent( xfs_alloc_arg_t *args) /* argument structure for allocation */ { int error=0; - xfs_extlen_t reservation; - xfs_extlen_t oldmax; ASSERT(args->minlen > 0); ASSERT(args->maxlen > 0); @@ -695,20 +669,6 @@ xfs_alloc_ag_vextent( ASSERT(args->mod < args->prod); ASSERT(args->alignment > 0); - /* - * Clamp maxlen to the amount of free space minus any reservations - * that have been made. - */ - oldmax = args->maxlen; - reservation = xfs_ag_resv_needed(args->pag, args->resv); - if (args->maxlen > args->pag->pagf_freeblks - reservation) - args->maxlen = args->pag->pagf_freeblks - reservation; - if (args->maxlen == 0) { - args->agbno = NULLAGBLOCK; - args->maxlen = oldmax; - return 0; - } - /* * Branch to correct routine based on the type. */ @@ -728,8 +688,6 @@ xfs_alloc_ag_vextent( /* NOTREACHED */ } - args->maxlen = oldmax; - if (error || args->agbno == NULLAGBLOCK) return error; @@ -838,9 +796,6 @@ xfs_alloc_ag_vextent_exact( args->len = XFS_AGBLOCK_MIN(tend, args->agbno + args->maxlen) - args->agbno; xfs_alloc_fix_len(args); - if (!xfs_alloc_fix_minleft(args)) - goto not_found; - ASSERT(args->agbno + args->len <= tend); /* @@ -1146,12 +1101,7 @@ restart: XFS_WANT_CORRUPTED_GOTO(args->mp, i == 1, error0); ASSERT(ltbno + ltlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length)); args->len = blen; - if (!xfs_alloc_fix_minleft(args)) { - xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); - trace_xfs_alloc_near_nominleft(args); - return 0; - } - blen = args->len; + /* * We are allocating starting at bnew for blen blocks. */ @@ -1343,12 +1293,6 @@ restart: */ args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen); xfs_alloc_fix_len(args); - if (!xfs_alloc_fix_minleft(args)) { - trace_xfs_alloc_near_nominleft(args); - xfs_btree_del_cursor(bno_cur_lt, XFS_BTREE_NOERROR); - xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); - return 0; - } rlen = args->len; (void)xfs_alloc_compute_diff(args->agbno, rlen, args->alignment, args->datatype, ltbnoa, ltlena, <new); @@ -1550,8 +1494,6 @@ restart: } xfs_alloc_fix_len(args); - if (!xfs_alloc_fix_minleft(args)) - goto out_nominleft; rlen = args->len; XFS_WANT_CORRUPTED_GOTO(args->mp, rlen <= flen, error0); /* @@ -2070,10 +2012,20 @@ xfs_alloc_space_available( /* do we have enough free space remaining for the allocation? */ available = (int)(pag->pagf_freeblks + pag->pagf_flcount - - reservation - min_free - args->total); - if (available < (int)args->minleft || available <= 0) + reservation - min_free - args->minleft); + if (available < (int)args->total) return false; + /* + * Clamp maxlen to the amount of free space available for the actual + * extent allocation. + */ + if (available < (int)args->maxlen && !(flags & XFS_ALLOC_FLAG_CHECK)) { + args->maxlen = available; + ASSERT(args->maxlen > 0); + ASSERT(args->maxlen >= args->minlen); + } + return true; } @@ -2119,7 +2071,8 @@ xfs_alloc_fix_freelist( } need = xfs_alloc_min_freelist(mp, pag); - if (!xfs_alloc_space_available(args, need, flags)) + if (!xfs_alloc_space_available(args, need, flags | + XFS_ALLOC_FLAG_CHECK)) goto out_agbp_relse; /* diff --git a/fs/xfs/libxfs/xfs_alloc.h b/fs/xfs/libxfs/xfs_alloc.h index 7c404a6b0ae3..1d0f48a501a3 100644 --- a/fs/xfs/libxfs/xfs_alloc.h +++ b/fs/xfs/libxfs/xfs_alloc.h @@ -56,7 +56,7 @@ typedef unsigned int xfs_alloctype_t; #define XFS_ALLOC_FLAG_FREEING 0x00000002 /* indicate caller is freeing extents*/ #define XFS_ALLOC_FLAG_NORMAP 0x00000004 /* don't modify the rmapbt */ #define XFS_ALLOC_FLAG_NOSHRINK 0x00000008 /* don't shrink the freelist */ - +#define XFS_ALLOC_FLAG_CHECK 0x00000010 /* test only, don't modify args */ /* * Argument structure for xfs_alloc routines. From e9b77651910722cd74a3936f230366503089bc80 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 2 Feb 2017 08:55:56 +0100 Subject: [PATCH 103/119] xfs: don't rely on ->total in xfs_alloc_space_available commit 12ef830198b0d71668eb9b59f9ba69d32951a48a upstream. ->total is a bit of an odd parameter passed down to the low-level allocator all the way from the high-level callers. It's supposed to contain the maximum number of blocks to be allocated for the whole transaction [1]. But in xfs_iomap_write_allocate we only convert existing delayed allocations and thus only have a minimal block reservation for the current transaction, so xfs_alloc_space_available can't use it for the allocation decisions. Use the maximum of args->total and the calculated block requirement to make a decision. We probably should get rid of args->total eventually and instead apply ->minleft more broadly, but that will require some extensive changes all over. [1] which creates lots of confusion as most callers don't decrement it once doing a first allocation. But that's for a separate series. Signed-off-by: Christoph Hellwig Reviewed-by: Brian Foster Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_alloc.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/fs/xfs/libxfs/xfs_alloc.c b/fs/xfs/libxfs/xfs_alloc.c index f2e7eb6e5243..9f06a211e157 100644 --- a/fs/xfs/libxfs/xfs_alloc.c +++ b/fs/xfs/libxfs/xfs_alloc.c @@ -1995,7 +1995,7 @@ xfs_alloc_space_available( int flags) { struct xfs_perag *pag = args->pag; - xfs_extlen_t longest; + xfs_extlen_t alloc_len, longest; xfs_extlen_t reservation; /* blocks that are still reserved */ int available; @@ -2005,15 +2005,16 @@ xfs_alloc_space_available( reservation = xfs_ag_resv_needed(pag, args->resv); /* do we have enough contiguous free space for the allocation? */ + alloc_len = args->minlen + (args->alignment - 1) + args->minalignslop; longest = xfs_alloc_longest_free_extent(args->mp, pag, min_free, reservation); - if ((args->minlen + args->alignment + args->minalignslop - 1) > longest) + if (longest < alloc_len) return false; /* do we have enough free space remaining for the allocation? */ available = (int)(pag->pagf_freeblks + pag->pagf_flcount - reservation - min_free - args->minleft); - if (available < (int)args->total) + if (available < (int)max(args->total, alloc_len)) return false; /* From 4f4d5082ec73f6e0bd44a8eccbef440b96f46a98 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 2 Feb 2017 08:55:57 +0100 Subject: [PATCH 104/119] xfs: don't print warnings when xfs_log_force fails commit 84a4620cfe97c9d57e39b2369bfb77faff55063d upstream. There are only two reasons for xfs_log_force / xfs_log_force_lsn to fail: one is an I/O error, for which xlog_bdstrat already logs a warning, and the second is an already shutdown log due to a previous I/O errors. In the latter case we'll already have a previous indication for the actual error, but the large stream of misleading warnings from xfs_log_force will probably scroll it out of the message buffer. Simply removing the warnings thus makes the XFS log reporting significantly better. Signed-off-by: Christoph Hellwig Reviewed-by: Carlos Maiolino Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/xfs_log.c | 12 ++---------- 1 file changed, 2 insertions(+), 10 deletions(-) diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 3b74fa011bb1..4017aa967331 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -3324,12 +3324,8 @@ xfs_log_force( xfs_mount_t *mp, uint flags) { - int error; - trace_xfs_log_force(mp, 0, _RET_IP_); - error = _xfs_log_force(mp, flags, NULL); - if (error) - xfs_warn(mp, "%s: error %d returned.", __func__, error); + _xfs_log_force(mp, flags, NULL); } /* @@ -3473,12 +3469,8 @@ xfs_log_force_lsn( xfs_lsn_t lsn, uint flags) { - int error; - trace_xfs_log_force(mp, lsn, _RET_IP_); - error = _xfs_log_force_lsn(mp, lsn, flags, NULL); - if (error) - xfs_warn(mp, "%s: error %d returned.", __func__, error); + _xfs_log_force_lsn(mp, lsn, flags, NULL); } /* From 624e54b5aff15c6d49974f404c5cef5ce7f89cd0 Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Thu, 2 Feb 2017 08:55:58 +0100 Subject: [PATCH 105/119] xfs: make the ASSERT() condition likely commit bf46ecc3d8cca05f2907cf482755c42c2b11a79d upstream. The ASSERT() condition is the normal case, not the exception, so testing the condition should be likely(), not unlikely(). Reviewed-by: Christoph Hellwig Signed-off-by: Amir Goldstein Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/xfs_linux.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/xfs/xfs_linux.h b/fs/xfs/xfs_linux.h index 68640fb63a54..1455b25205a8 100644 --- a/fs/xfs/xfs_linux.h +++ b/fs/xfs/xfs_linux.h @@ -330,11 +330,11 @@ static inline __uint64_t howmany_64(__uint64_t x, __uint32_t y) } #define ASSERT_ALWAYS(expr) \ - (unlikely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__)) + (likely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__)) #ifdef DEBUG #define ASSERT(expr) \ - (unlikely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__)) + (likely(expr) ? (void)0 : assfail(#expr, __FILE__, __LINE__)) #ifndef STATIC # define STATIC noinline @@ -345,7 +345,7 @@ static inline __uint64_t howmany_64(__uint64_t x, __uint32_t y) #ifdef XFS_WARN #define ASSERT(expr) \ - (unlikely(expr) ? (void)0 : asswarn(#expr, __FILE__, __LINE__)) + (likely(expr) ? (void)0 : asswarn(#expr, __FILE__, __LINE__)) #ifndef STATIC # define STATIC static noinline From e5325fcf70b1cd2c45503b854fde3f1bd44216da Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Thu, 2 Feb 2017 08:55:59 +0100 Subject: [PATCH 106/119] xfs: sanity check directory inode di_size commit 3c6f46eacd876bd723a9bad3c6882714c052fd8e upstream. This changes fixes an assertion hit when fuzzing on-disk i_mode values. The easy case to fix is when changing an empty file i_mode to S_IFDIR. In this case, xfs_dinode_verify() detects an illegal zero size for directory and fails to load the inode structure from disk. For the case of non empty file whose i_mode is changed to S_IFDIR, the ASSERT() statement in xfs_dir2_isblock() is replaced with return -EFSCORRUPTED, to avoid interacting with corrupted jusk also when XFS_DEBUG is disabled. Suggested-by: Darrick J. Wong Reviewed-by: Christoph Hellwig Signed-off-by: Amir Goldstein Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_dir2.c | 3 ++- fs/xfs/libxfs/xfs_inode_buf.c | 7 +++++-- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/fs/xfs/libxfs/xfs_dir2.c b/fs/xfs/libxfs/xfs_dir2.c index 20a96dd5af7e..ec326d272efb 100644 --- a/fs/xfs/libxfs/xfs_dir2.c +++ b/fs/xfs/libxfs/xfs_dir2.c @@ -631,7 +631,8 @@ xfs_dir2_isblock( if ((rval = xfs_bmap_last_offset(args->dp, &last, XFS_DATA_FORK))) return rval; rval = XFS_FSB_TO_B(args->dp->i_mount, last) == args->geo->blksize; - ASSERT(rval == 0 || args->dp->i_d.di_size == args->geo->blksize); + if (rval != 0 && args->dp->i_d.di_size != args->geo->blksize) + return -EFSCORRUPTED; *vp = rval; return 0; } diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c index c906e50515f0..977245de8953 100644 --- a/fs/xfs/libxfs/xfs_inode_buf.c +++ b/fs/xfs/libxfs/xfs_inode_buf.c @@ -386,6 +386,7 @@ xfs_dinode_verify( struct xfs_inode *ip, struct xfs_dinode *dip) { + uint16_t mode; uint16_t flags; uint64_t flags2; @@ -396,8 +397,10 @@ xfs_dinode_verify( if (be64_to_cpu(dip->di_size) & (1ULL << 63)) return false; - /* No zero-length symlinks. */ - if (S_ISLNK(be16_to_cpu(dip->di_mode)) && dip->di_size == 0) + mode = be16_to_cpu(dip->di_mode); + + /* No zero-length symlinks/dirs. */ + if ((S_ISLNK(mode) || S_ISDIR(mode)) && dip->di_size == 0) return false; /* only version 3 or greater inodes are extensively verified here */ From 4fac84ba1da7aa62dea520dcedd4f6de117d8f2b Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Thu, 2 Feb 2017 08:56:00 +0100 Subject: [PATCH 107/119] xfs: add missing include dependencies to xfs_dir2.h commit b597dd5373a1ccc08218665dc8417433b1c09550 upstream. xfs_dir2.h dereferences some data types in inline functions and fails to include those type definitions, e.g.: xfs_dir2_data_aoff_t, struct xfs_da_geometry. Signed-off-by: Amir Goldstein Reviewed-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_dir2.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/xfs/libxfs/xfs_dir2.h b/fs/xfs/libxfs/xfs_dir2.h index becc926c3e3d..6a3fe31d4bc2 100644 --- a/fs/xfs/libxfs/xfs_dir2.h +++ b/fs/xfs/libxfs/xfs_dir2.h @@ -18,6 +18,9 @@ #ifndef __XFS_DIR2_H__ #define __XFS_DIR2_H__ +#include "xfs_da_format.h" +#include "xfs_da_btree.h" + struct xfs_defer_ops; struct xfs_da_args; struct xfs_inode; From b5f68e24cc7bc3492ebc5c70f3ef6babcbd4188b Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Thu, 2 Feb 2017 08:56:01 +0100 Subject: [PATCH 108/119] xfs: replace xfs_mode_to_ftype table with switch statement commit 1fc4d33fed124fb182e8e6c214e973a29389ae83. The size of the xfs_mode_to_ftype[] conversion table was too small to handle an invalid value of mode=S_IFMT. Instead of fixing the table size, replace the conversion table with a conversion helper that uses a switch statement. Suggested-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Reviewed-by: Christoph Hellwig Signed-off-by: Amir Goldstein Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_dir2.c | 36 ++++++++++++++++++++++-------------- fs/xfs/libxfs/xfs_dir2.h | 5 ++--- fs/xfs/xfs_iops.c | 2 +- 3 files changed, 25 insertions(+), 18 deletions(-) diff --git a/fs/xfs/libxfs/xfs_dir2.c b/fs/xfs/libxfs/xfs_dir2.c index ec326d272efb..1a978bd9d506 100644 --- a/fs/xfs/libxfs/xfs_dir2.c +++ b/fs/xfs/libxfs/xfs_dir2.c @@ -36,21 +36,29 @@ struct xfs_name xfs_name_dotdot = { (unsigned char *)"..", 2, XFS_DIR3_FT_DIR }; /* - * @mode, if set, indicates that the type field needs to be set up. - * This uses the transformation from file mode to DT_* as defined in linux/fs.h - * for file type specification. This will be propagated into the directory - * structure if appropriate for the given operation and filesystem config. + * Convert inode mode to directory entry filetype */ -const unsigned char xfs_mode_to_ftype[S_IFMT >> S_SHIFT] = { - [0] = XFS_DIR3_FT_UNKNOWN, - [S_IFREG >> S_SHIFT] = XFS_DIR3_FT_REG_FILE, - [S_IFDIR >> S_SHIFT] = XFS_DIR3_FT_DIR, - [S_IFCHR >> S_SHIFT] = XFS_DIR3_FT_CHRDEV, - [S_IFBLK >> S_SHIFT] = XFS_DIR3_FT_BLKDEV, - [S_IFIFO >> S_SHIFT] = XFS_DIR3_FT_FIFO, - [S_IFSOCK >> S_SHIFT] = XFS_DIR3_FT_SOCK, - [S_IFLNK >> S_SHIFT] = XFS_DIR3_FT_SYMLINK, -}; +const unsigned char xfs_mode_to_ftype(int mode) +{ + switch (mode & S_IFMT) { + case S_IFREG: + return XFS_DIR3_FT_REG_FILE; + case S_IFDIR: + return XFS_DIR3_FT_DIR; + case S_IFCHR: + return XFS_DIR3_FT_CHRDEV; + case S_IFBLK: + return XFS_DIR3_FT_BLKDEV; + case S_IFIFO: + return XFS_DIR3_FT_FIFO; + case S_IFSOCK: + return XFS_DIR3_FT_SOCK; + case S_IFLNK: + return XFS_DIR3_FT_SYMLINK; + default: + return XFS_DIR3_FT_UNKNOWN; + } +} /* * ASCII case-insensitive (ie. A-Z) support for directories that was diff --git a/fs/xfs/libxfs/xfs_dir2.h b/fs/xfs/libxfs/xfs_dir2.h index 6a3fe31d4bc2..0051a34b8c3f 100644 --- a/fs/xfs/libxfs/xfs_dir2.h +++ b/fs/xfs/libxfs/xfs_dir2.h @@ -35,10 +35,9 @@ struct xfs_dir2_data_unused; extern struct xfs_name xfs_name_dotdot; /* - * directory filetype conversion tables. + * Convert inode mode to directory entry filetype */ -#define S_SHIFT 12 -extern const unsigned char xfs_mode_to_ftype[]; +extern const unsigned char xfs_mode_to_ftype(int mode); /* * directory operations vector for encode/decode routines diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index 405a65cd9d6b..1abe71918734 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -103,7 +103,7 @@ xfs_dentry_to_name( { namep->name = dentry->d_name.name; namep->len = dentry->d_name.len; - namep->type = xfs_mode_to_ftype[(mode & S_IFMT) >> S_SHIFT]; + namep->type = xfs_mode_to_ftype(mode); } STATIC void From 43ce5921773b827531ed123016625481d43c2bd9 Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Thu, 2 Feb 2017 08:56:02 +0100 Subject: [PATCH 109/119] xfs: sanity check inode mode when creating new dentry commit fab8eef86c814c3dd46bc5d760b6e4a53d5fc5a6 upstream. The helper xfs_dentry_to_name() is used by 2 different classes of callers: Callers that pass zero mode and don't care about the returned name.type field and Callers that pass non zero mode and do care about the name.type field. Change xfs_dentry_to_name() to not take the mode argument and change the call sites of the first class to not pass the mode argument. Create a new helper xfs_dentry_mode_to_name() which does pass the mode argument and returns -EFSCORRUPTED if mode is invalid. Callers that translate non zero mode to on-disk file type now check the return value and will export the error to user instead of staging an invalid file type to be written to directory entry. Signed-off-by: Amir Goldstein Reviewed-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/xfs_iops.c | 48 ++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 39 insertions(+), 9 deletions(-) diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index 1abe71918734..f5e0f608e245 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -97,6 +97,16 @@ xfs_init_security( static void xfs_dentry_to_name( + struct xfs_name *namep, + struct dentry *dentry) +{ + namep->name = dentry->d_name.name; + namep->len = dentry->d_name.len; + namep->type = XFS_DIR3_FT_UNKNOWN; +} + +static int +xfs_dentry_mode_to_name( struct xfs_name *namep, struct dentry *dentry, int mode) @@ -104,6 +114,11 @@ xfs_dentry_to_name( namep->name = dentry->d_name.name; namep->len = dentry->d_name.len; namep->type = xfs_mode_to_ftype(mode); + + if (unlikely(namep->type == XFS_DIR3_FT_UNKNOWN)) + return -EFSCORRUPTED; + + return 0; } STATIC void @@ -119,7 +134,7 @@ xfs_cleanup_inode( * xfs_init_security we must back out. * ENOSPC can hit here, among other things. */ - xfs_dentry_to_name(&teardown, dentry, 0); + xfs_dentry_to_name(&teardown, dentry); xfs_remove(XFS_I(dir), &teardown, XFS_I(inode)); } @@ -154,8 +169,12 @@ xfs_generic_create( if (error) return error; + /* Verify mode is valid also for tmpfile case */ + error = xfs_dentry_mode_to_name(&name, dentry, mode); + if (unlikely(error)) + goto out_free_acl; + if (!tmpfile) { - xfs_dentry_to_name(&name, dentry, mode); error = xfs_create(XFS_I(dir), &name, mode, rdev, &ip); } else { error = xfs_create_tmpfile(XFS_I(dir), dentry, mode, &ip); @@ -248,7 +267,7 @@ xfs_vn_lookup( if (dentry->d_name.len >= MAXNAMELEN) return ERR_PTR(-ENAMETOOLONG); - xfs_dentry_to_name(&name, dentry, 0); + xfs_dentry_to_name(&name, dentry); error = xfs_lookup(XFS_I(dir), &name, &cip, NULL); if (unlikely(error)) { if (unlikely(error != -ENOENT)) @@ -275,7 +294,7 @@ xfs_vn_ci_lookup( if (dentry->d_name.len >= MAXNAMELEN) return ERR_PTR(-ENAMETOOLONG); - xfs_dentry_to_name(&xname, dentry, 0); + xfs_dentry_to_name(&xname, dentry); error = xfs_lookup(XFS_I(dir), &xname, &ip, &ci_name); if (unlikely(error)) { if (unlikely(error != -ENOENT)) @@ -310,7 +329,9 @@ xfs_vn_link( struct xfs_name name; int error; - xfs_dentry_to_name(&name, dentry, inode->i_mode); + error = xfs_dentry_mode_to_name(&name, dentry, inode->i_mode); + if (unlikely(error)) + return error; error = xfs_link(XFS_I(dir), XFS_I(inode), &name); if (unlikely(error)) @@ -329,7 +350,7 @@ xfs_vn_unlink( struct xfs_name name; int error; - xfs_dentry_to_name(&name, dentry, 0); + xfs_dentry_to_name(&name, dentry); error = xfs_remove(XFS_I(dir), &name, XFS_I(d_inode(dentry))); if (error) @@ -359,7 +380,9 @@ xfs_vn_symlink( mode = S_IFLNK | (irix_symlink_mode ? 0777 & ~current_umask() : S_IRWXUGO); - xfs_dentry_to_name(&name, dentry, mode); + error = xfs_dentry_mode_to_name(&name, dentry, mode); + if (unlikely(error)) + goto out; error = xfs_symlink(XFS_I(dir), &name, symname, mode, &cip); if (unlikely(error)) @@ -395,6 +418,7 @@ xfs_vn_rename( { struct inode *new_inode = d_inode(ndentry); int omode = 0; + int error; struct xfs_name oname; struct xfs_name nname; @@ -405,8 +429,14 @@ xfs_vn_rename( if (flags & RENAME_EXCHANGE) omode = d_inode(ndentry)->i_mode; - xfs_dentry_to_name(&oname, odentry, omode); - xfs_dentry_to_name(&nname, ndentry, d_inode(odentry)->i_mode); + error = xfs_dentry_mode_to_name(&oname, odentry, omode); + if (omode && unlikely(error)) + return error; + + error = xfs_dentry_mode_to_name(&nname, ndentry, + d_inode(odentry)->i_mode); + if (unlikely(error)) + return error; return xfs_rename(XFS_I(odir), &oname, XFS_I(d_inode(odentry)), XFS_I(ndir), &nname, From d3201a14b36ae6ffb36b6b830e216f0ae4ebc184 Mon Sep 17 00:00:00 2001 From: Amir Goldstein Date: Thu, 2 Feb 2017 08:56:03 +0100 Subject: [PATCH 110/119] xfs: sanity check inode di_mode commit a324cbf10a3c67aaa10c9f47f7b5801562925bc2 upstream. Check for invalid file type in xfs_dinode_verify() and fail to load the inode structure from disk. Reviewed-by: Darrick J. Wong Signed-off-by: Amir Goldstein Reviewed-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_inode_buf.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/xfs/libxfs/xfs_inode_buf.c b/fs/xfs/libxfs/xfs_inode_buf.c index 977245de8953..37ee7f01a35d 100644 --- a/fs/xfs/libxfs/xfs_inode_buf.c +++ b/fs/xfs/libxfs/xfs_inode_buf.c @@ -29,6 +29,7 @@ #include "xfs_icache.h" #include "xfs_trans.h" #include "xfs_ialloc.h" +#include "xfs_dir2.h" /* * Check that none of the inode's in the buffer have a next @@ -398,6 +399,8 @@ xfs_dinode_verify( return false; mode = be16_to_cpu(dip->di_mode); + if (mode && xfs_mode_to_ftype(mode) == XFS_DIR3_FT_UNKNOWN) + return false; /* No zero-length symlinks/dirs. */ if ((S_ISLNK(mode) || S_ISDIR(mode)) && dip->di_size == 0) From d062d90c35f292b3e6ee266746259e2e2950f940 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 2 Feb 2017 08:56:04 +0100 Subject: [PATCH 111/119] xfs: don't wrap ID in xfs_dq_get_next_id commit 657bdfb7f5e68ca5e2ed009ab473c429b0d6af85 upstream. The GETNEXTQOTA ioctl takes whatever ID is sent in, and looks for the next active quota for an user equal or higher to that ID. But if we are at the maximum ID and then ask for the "next" one, we may wrap back to zero. In this case, userspace may loop forever, because it will start querying again at zero. We'll fix this in userspace as well, but for the kernel, return -ENOENT if we ask for the next quota ID past UINT_MAX so the caller knows to stop. Signed-off-by: Eric Sandeen Reviewed-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/xfs_dquot.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c index 7a30b8f11db7..9d06cc30e875 100644 --- a/fs/xfs/xfs_dquot.c +++ b/fs/xfs/xfs_dquot.c @@ -710,6 +710,10 @@ xfs_dq_get_next_id( /* Simple advance */ next_id = *id + 1; + /* If we'd wrap past the max ID, stop */ + if (next_id < *id) + return -ENOENT; + /* If new ID is within the current chunk, advancing it sufficed */ if (next_id % mp->m_quotainfo->qi_dqperchunk) { *id = next_id; From 29f319275e7637b3d146aa67db7fad036339fcc1 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Thu, 2 Feb 2017 08:56:05 +0100 Subject: [PATCH 112/119] xfs: fix xfs_mode_to_ftype() prototype commit fd29f7af75b7adf250beccffa63746c6a88e2b74 upstream. A harmless warning just got introduced: fs/xfs/libxfs/xfs_dir2.h:40:8: error: type qualifiers ignored on function return type [-Werror=ignored-qualifiers] Removing the 'const' modifier avoids the warning and has no other effect. Fixes: 1fc4d33fed12 ("xfs: replace xfs_mode_to_ftype table with switch statement") Signed-off-by: Arnd Bergmann Reviewed-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_dir2.c | 2 +- fs/xfs/libxfs/xfs_dir2.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/xfs/libxfs/xfs_dir2.c b/fs/xfs/libxfs/xfs_dir2.c index 1a978bd9d506..7825d78d4587 100644 --- a/fs/xfs/libxfs/xfs_dir2.c +++ b/fs/xfs/libxfs/xfs_dir2.c @@ -38,7 +38,7 @@ struct xfs_name xfs_name_dotdot = { (unsigned char *)"..", 2, XFS_DIR3_FT_DIR }; /* * Convert inode mode to directory entry filetype */ -const unsigned char xfs_mode_to_ftype(int mode) +unsigned char xfs_mode_to_ftype(int mode) { switch (mode & S_IFMT) { case S_IFREG: diff --git a/fs/xfs/libxfs/xfs_dir2.h b/fs/xfs/libxfs/xfs_dir2.h index 0051a34b8c3f..ae0d55bf6500 100644 --- a/fs/xfs/libxfs/xfs_dir2.h +++ b/fs/xfs/libxfs/xfs_dir2.h @@ -37,7 +37,7 @@ extern struct xfs_name xfs_name_dotdot; /* * Convert inode mode to directory entry filetype */ -extern const unsigned char xfs_mode_to_ftype(int mode); +extern unsigned char xfs_mode_to_ftype(int mode); /* * directory operations vector for encode/decode routines From 214d55efa25557ca4d023c4001d798b7d360cd8b Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 2 Feb 2017 08:56:06 +0100 Subject: [PATCH 113/119] xfs: fix COW writeback race commit d2b3964a0780d2d2994eba57f950d6c9fe489ed8 upstream. Due to the way how xfs_iomap_write_allocate tries to convert the whole found extents from delalloc to real space we can run into a race condition with multiple threads doing writes to this same extent. For the non-COW case that is harmless as the only thing that can happen is that we call xfs_bmapi_write on an extent that has already been converted to a real allocation. For COW writes where we move the extent from the COW to the data fork after I/O completion the race is, however, not quite as harmless. In the worst case we are now calling xfs_bmapi_write on a region that contains hole in the COW work, which will trip up an assert in debug builds or lead to file system corruption in non-debug builds. This seems to be reproducible with workloads of small O_DSYNC write, although so far I've not managed to come up with a with an isolated reproducer. The fix for the issue is relatively simple: tell xfs_bmapi_write that we are only asked to convert delayed allocations and skip holes in that case. Signed-off-by: Christoph Hellwig Reviewed-by: Brian Foster Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_bmap.c | 44 +++++++++++++++++++++++++++++----------- fs/xfs/libxfs/xfs_bmap.h | 6 +++++- fs/xfs/xfs_iomap.c | 2 +- 3 files changed, 38 insertions(+), 14 deletions(-) diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index 52dc5c175001..fbb60d30089c 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -4607,8 +4607,6 @@ xfs_bmapi_write( int n; /* current extent index */ xfs_fileoff_t obno; /* old block number (offset) */ int whichfork; /* data or attr fork */ - char inhole; /* current location is hole in file */ - char wasdelay; /* old extent was delayed */ #ifdef DEBUG xfs_fileoff_t orig_bno; /* original block number value */ @@ -4694,22 +4692,44 @@ xfs_bmapi_write( bma.firstblock = firstblock; while (bno < end && n < *nmap) { - inhole = eof || bma.got.br_startoff > bno; - wasdelay = !inhole && isnullstartblock(bma.got.br_startblock); + bool need_alloc = false, wasdelay = false; - /* - * Make sure we only reflink into a hole. - */ - if (flags & XFS_BMAPI_REMAP) - ASSERT(inhole); - if (flags & XFS_BMAPI_COWFORK) - ASSERT(!inhole); + /* in hole or beyoned EOF? */ + if (eof || bma.got.br_startoff > bno) { + if (flags & XFS_BMAPI_DELALLOC) { + /* + * For the COW fork we can reasonably get a + * request for converting an extent that races + * with other threads already having converted + * part of it, as there converting COW to + * regular blocks is not protected using the + * IOLOCK. + */ + ASSERT(flags & XFS_BMAPI_COWFORK); + if (!(flags & XFS_BMAPI_COWFORK)) { + error = -EIO; + goto error0; + } + + if (eof || bno >= end) + break; + } else { + need_alloc = true; + } + } else { + /* + * Make sure we only reflink into a hole. + */ + ASSERT(!(flags & XFS_BMAPI_REMAP)); + if (isnullstartblock(bma.got.br_startblock)) + wasdelay = true; + } /* * First, deal with the hole before the allocated space * that we found, if any. */ - if (inhole || wasdelay) { + if (need_alloc || wasdelay) { bma.eof = eof; bma.conv = !!(flags & XFS_BMAPI_CONVERT); bma.wasdel = wasdelay; diff --git a/fs/xfs/libxfs/xfs_bmap.h b/fs/xfs/libxfs/xfs_bmap.h index d6d175a4fdec..e7d40b39f18f 100644 --- a/fs/xfs/libxfs/xfs_bmap.h +++ b/fs/xfs/libxfs/xfs_bmap.h @@ -110,6 +110,9 @@ struct xfs_extent_free_item /* Map something in the CoW fork. */ #define XFS_BMAPI_COWFORK 0x200 +/* Only convert delalloc space, don't allocate entirely new extents */ +#define XFS_BMAPI_DELALLOC 0x400 + #define XFS_BMAPI_FLAGS \ { XFS_BMAPI_ENTIRE, "ENTIRE" }, \ { XFS_BMAPI_METADATA, "METADATA" }, \ @@ -120,7 +123,8 @@ struct xfs_extent_free_item { XFS_BMAPI_CONVERT, "CONVERT" }, \ { XFS_BMAPI_ZERO, "ZERO" }, \ { XFS_BMAPI_REMAP, "REMAP" }, \ - { XFS_BMAPI_COWFORK, "COWFORK" } + { XFS_BMAPI_COWFORK, "COWFORK" }, \ + { XFS_BMAPI_DELALLOC, "DELALLOC" } static inline int xfs_bmapi_aflag(int w) diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 15a83813b708..cdc6bdd495be 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -681,7 +681,7 @@ xfs_iomap_write_allocate( xfs_trans_t *tp; int nimaps; int error = 0; - int flags = 0; + int flags = XFS_BMAPI_DELALLOC; int nres; if (whichfork == XFS_COW_FORK) From 29094164ea999bfc48ef48780c1ae057afaafcb1 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 2 Feb 2017 08:56:07 +0100 Subject: [PATCH 114/119] xfs: verify dirblocklog correctly commit 83d230eb5c638949350f4761acdfc0af5cb1bc00 upstream. sb_dirblklog is added to sb_blocklog to compute the directory block size in bytes. Therefore, we must compare the sum of both those values against XFS_MAX_BLOCKSIZE_LOG, not just dirblklog. Signed-off-by: Darrick J. Wong Reviewed-by: Eric Sandeen Reviewed-by: Christoph Hellwig Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_sb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/libxfs/xfs_sb.c b/fs/xfs/libxfs/xfs_sb.c index 2580262e4ea0..584ec896a533 100644 --- a/fs/xfs/libxfs/xfs_sb.c +++ b/fs/xfs/libxfs/xfs_sb.c @@ -242,7 +242,7 @@ xfs_mount_validate_sb( sbp->sb_blocklog < XFS_MIN_BLOCKSIZE_LOG || sbp->sb_blocklog > XFS_MAX_BLOCKSIZE_LOG || sbp->sb_blocksize != (1 << sbp->sb_blocklog) || - sbp->sb_dirblklog > XFS_MAX_BLOCKSIZE_LOG || + sbp->sb_dirblklog + sbp->sb_blocklog > XFS_MAX_BLOCKSIZE_LOG || sbp->sb_inodesize < XFS_DINODE_MIN_SIZE || sbp->sb_inodesize > XFS_DINODE_MAX_SIZE || sbp->sb_inodelog < XFS_DINODE_MIN_LOG || From aab858dabd5ee4df1d8876449db2ef7868d1b5d0 Mon Sep 17 00:00:00 2001 From: Brian Foster Date: Thu, 2 Feb 2017 08:56:08 +0100 Subject: [PATCH 115/119] xfs: remove racy hasattr check from attr ops commit 5a93790d4e2df73e30c965ec6e49be82fc3ccfce upstream. xfs_attr_[get|remove]() have unlocked attribute fork checks to optimize away a lock cycle in cases where the fork does not exist or is otherwise empty. This check is not safe, however, because an attribute fork short form to extent format conversion includes a transient state that causes the xfs_inode_hasattr() check to fail. Specifically, xfs_attr_shortform_to_leaf() creates an empty extent format attribute fork and then adds the existing shortform attributes to it. This means that lookup of an existing xattr can spuriously return -ENOATTR when racing against a setxattr that causes the associated format conversion. This was originally reproduced by an untar on a particularly configured glusterfs volume, but can also be reproduced on demand with properly crafted xattr requests. The format conversion occurs under the exclusive ilock. xfs_attr_get() and xfs_attr_remove() already have the proper locking and checks further down in the functions to handle this situation correctly. Drop the unlocked checks to avoid the spurious failure and rely on the existing logic. Signed-off-by: Brian Foster Reviewed-by: Christoph Hellwig Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_attr.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/fs/xfs/libxfs/xfs_attr.c b/fs/xfs/libxfs/xfs_attr.c index af1ecb19121e..6622d46ddec3 100644 --- a/fs/xfs/libxfs/xfs_attr.c +++ b/fs/xfs/libxfs/xfs_attr.c @@ -131,9 +131,6 @@ xfs_attr_get( if (XFS_FORCED_SHUTDOWN(ip->i_mount)) return -EIO; - if (!xfs_inode_hasattr(ip)) - return -ENOATTR; - error = xfs_attr_args_init(&args, ip, name, flags); if (error) return error; @@ -392,9 +389,6 @@ xfs_attr_remove( if (XFS_FORCED_SHUTDOWN(dp->i_mount)) return -EIO; - if (!xfs_inode_hasattr(dp)) - return -ENOATTR; - error = xfs_attr_args_init(&args, dp, name, flags); if (error) return error; From 29f96b7e9023929f8bd80b5e1f53d4e6db3c434f Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 2 Feb 2017 08:56:09 +0100 Subject: [PATCH 116/119] xfs: extsize hints are not unlikely in xfs_bmap_btalloc commit 493611ebd62673f39e2f52c2561182c558a21cb6 upstream. With COW files they are the hotpath, just like for files with the extent size hint attribute. We really shouldn't micro-manage anything but failure cases with unlikely. Additionally Arnd Bergmann recently reported that one of these two unlikely annotations causes link failures together with an upcoming kernel instrumentation patch, so let's get rid of it ASAP. Signed-off-by: Christoph Hellwig Reported-by: Arnd Bergmann Reviewed-by: Darrick J. Wong Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/libxfs/xfs_bmap.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c index fbb60d30089c..f52fd63fce19 100644 --- a/fs/xfs/libxfs/xfs_bmap.c +++ b/fs/xfs/libxfs/xfs_bmap.c @@ -3720,7 +3720,7 @@ xfs_bmap_btalloc( align = xfs_get_cowextsz_hint(ap->ip); else if (xfs_alloc_is_userdata(ap->datatype)) align = xfs_get_extsz_hint(ap->ip); - if (unlikely(align)) { + if (align) { error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev, align, 0, ap->eof, 0, ap->conv, &ap->offset, &ap->length); @@ -3792,7 +3792,7 @@ xfs_bmap_btalloc( args.minlen = ap->minlen; } /* apply extent size hints if obtained earlier */ - if (unlikely(align)) { + if (align) { args.prod = align; if ((args.mod = (xfs_extlen_t)do_mod(ap->offset, args.prod))) args.mod = (xfs_extlen_t)(args.prod - args.mod); From 5d44dd54bd57c6275d82d8912730c794fc8ec8ab Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 2 Feb 2017 08:56:10 +0100 Subject: [PATCH 117/119] xfs: clear _XBF_PAGES from buffers when readahead page commit 2aa6ba7b5ad3189cc27f14540aa2f57f0ed8df4b upstream. If we try to allocate memory pages to back an xfs_buf that we're trying to read, it's possible that we'll be so short on memory that the page allocation fails. For a blocking read we'll just wait, but for readahead we simply dump all the pages we've collected so far. Unfortunately, after dumping the pages we neglect to clear the _XBF_PAGES state, which means that the subsequent call to xfs_buf_free thinks that b_pages still points to pages we own. It then double-frees the b_pages pages. This results in screaming about negative page refcounts from the memory manager, which xfs oughtn't be triggering. To reproduce this case, mount a filesystem where the size of the inodes far outweighs the availalble memory (a ~500M inode filesystem on a VM with 300MB memory did the trick here) and run bulkstat in parallel with other memory eating processes to put a huge load on the system. The "check summary" phase of xfs_scrub also works for this purpose. Signed-off-by: Darrick J. Wong Reviewed-by: Eric Sandeen Signed-off-by: Greg Kroah-Hartman --- fs/xfs/xfs_buf.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index b5b9bffe3520..d7a67d7fbc7f 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -423,6 +423,7 @@ retry: out_free_pages: for (i = 0; i < bp->b_page_count; i++) __free_page(bp->b_pages[i]); + bp->b_flags &= ~_XBF_PAGES; return error; } From b5b4d4a9141e15ea8d887d88d9763cf190955907 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Thu, 2 Feb 2017 08:56:11 +0100 Subject: [PATCH 118/119] xfs: fix bmv_count confusion w/ shared extents commit c364b6d0b6cda1cd5d9ab689489adda3e82529aa upstream. In a bmapx call, bmv_count is the total size of the array, including the zeroth element that userspace uses to supply the search key. The output array starts at offset 1 so that we can set up the user for the next invocation. Since we now can split an extent into multiple bmap records due to shared/unshared status, we have to be careful that we don't overflow the output array. In the original patch f86f403794b ("xfs: teach get_bmapx about shared extents and the CoW fork") I used cur_ext (the output index) to check for overflows, albeit with an off-by-one error. Since nexleft no longer describes the number of unfilled slots in the output, we can rip all that out and use cur_ext for the overflow check directly. Failure to do this causes heap corruption in bmapx callers such as xfs_io and xfs_scrub. xfs/328 can reproduce this problem. Reviewed-by: Eric Sandeen Signed-off-by: Darrick J. Wong Signed-off-by: Greg Kroah-Hartman --- fs/xfs/xfs_bmap_util.c | 28 ++++++++++++++++++---------- 1 file changed, 18 insertions(+), 10 deletions(-) diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c index 0670a8bd5818..efb8ccd6bbf2 100644 --- a/fs/xfs/xfs_bmap_util.c +++ b/fs/xfs/xfs_bmap_util.c @@ -528,7 +528,6 @@ xfs_getbmap( xfs_bmbt_irec_t *map; /* buffer for user's data */ xfs_mount_t *mp; /* file system mount point */ int nex; /* # of user extents can do */ - int nexleft; /* # of user extents left */ int subnex; /* # of bmapi's can do */ int nmap; /* number of map entries */ struct getbmapx *out; /* output structure */ @@ -686,10 +685,8 @@ xfs_getbmap( goto out_free_map; } - nexleft = nex; - do { - nmap = (nexleft > subnex) ? subnex : nexleft; + nmap = (nex> subnex) ? subnex : nex; error = xfs_bmapi_read(ip, XFS_BB_TO_FSBT(mp, bmv->bmv_offset), XFS_BB_TO_FSB(mp, bmv->bmv_length), map, &nmap, bmapi_flags); @@ -697,8 +694,8 @@ xfs_getbmap( goto out_free_map; ASSERT(nmap <= subnex); - for (i = 0; i < nmap && nexleft && bmv->bmv_length && - cur_ext < bmv->bmv_count; i++) { + for (i = 0; i < nmap && bmv->bmv_length && + cur_ext < bmv->bmv_count - 1; i++) { out[cur_ext].bmv_oflags = 0; if (map[i].br_state == XFS_EXT_UNWRITTEN) out[cur_ext].bmv_oflags |= BMV_OF_PREALLOC; @@ -760,16 +757,27 @@ xfs_getbmap( continue; } + /* + * In order to report shared extents accurately, + * we report each distinct shared/unshared part + * of a single bmbt record using multiple bmap + * extents. To make that happen, we iterate the + * same map array item multiple times, each + * time trimming out the subextent that we just + * reported. + * + * Because of this, we must check the out array + * index (cur_ext) directly against bmv_count-1 + * to avoid overflows. + */ if (inject_map.br_startblock != NULLFSBLOCK) { map[i] = inject_map; i--; - } else - nexleft--; + } bmv->bmv_entries++; cur_ext++; } - } while (nmap && nexleft && bmv->bmv_length && - cur_ext < bmv->bmv_count); + } while (nmap && bmv->bmv_length && cur_ext < bmv->bmv_count - 1); out_free_map: kmem_free(map); From c8ea2f3b8247b6046f02527127ecf2fd0b045f46 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Sat, 4 Feb 2017 09:47:29 +0100 Subject: [PATCH 119/119] Linux 4.9.8 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index da704d903321..1130803ab93c 100644 --- a/Makefile +++ b/Makefile @@ -1,6 +1,6 @@ VERSION = 4 PATCHLEVEL = 9 -SUBLEVEL = 7 +SUBLEVEL = 8 EXTRAVERSION = NAME = Roaring Lionus