This commit is now upstream in 5.17-rc1, along with a number of other
changes and cleanups on top of it, so it should be removed from the tree
now to make merges easier.
Fixes: 60500a4228 ("ANDROID: mm: add a field to store names for private anonymous memory")
Cc: Colin Cross <ccross@android.com>
Cc: Dmitry Shmidt <dimitrysh@google.com>
Cc: Amit Pundir <amit.pundir@linaro.org>
Cc: Eric Biggers <ebiggers@google.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I4ceab971293adc9409636b8ccfea08f66b5b6d6b
8250_core registers 4 ISA uart ports by default, which can cause
problems on some devices which don't have them. This change doesn't
break earlycon=uart8250, but it will cause the 8250_of and 8250_pci sub
drivers to be unable to register ports. Boards that really need the full
8250 driver to take over from earlycon can use the "8250.nr_uarts=X"
kernel command line option to restore the ports allocation.
Bug: 216312411
Signed-off-by: Alistair Delva <adelva@google.com>
Change-Id: I04715394b32bd98544657101de4537df34554ea9
Steps on the way to 5.17-rc1
Resolves merge conflicts with:
mm/memremap.c
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I506780b8aa65e17a84fd9adb748e16825a3f4833
This reverts commit 4ba1429270.
When BTF generation is enabled for modules, we can have module load
failures due to BTF mismatches even when ABI is unchanged. Currently
Kconfig does not allow disabling BTF for modules only, so turn it off
entirely for now.
Signed-off-by: Connor O'Brien <connoro@google.com>
Change-Id: I0a7251397f95d6c02ba999bcf2ab9377ac0d76c3
This reverts commit 489a71964f.
Rationale does not apply to Android userspace which does not allow
access to debugfs on user builds.
Bug: 218319878
Signed-off-by: Alistair Delva <adelva@google.com>
Change-Id: Ia36b04b432a3db330e5a12763a3250869b97f0fe
Signed-off-by: Steve Muckle <smuckle@google.com>
blake2s_compress_generic is weakly aliased by blake2s_compress. The
current harness for function selection uses a function pointer, which is
ordinarily inlined and resolved at compile time. But when Clang's CFI is
enabled, CFI still triggers when making an indirect call via a weak
symbol. This seems like a bug in Clang's CFI, as though it's bucketing
weak symbols and strong symbols differently. It also only seems to
trigger when "full LTO" mode is used, rather than "thin LTO".
[ 0.000000][ T0] Kernel panic - not syncing: CFI failure (target: blake2s_compress_generic+0x0/0x1444)
[ 0.000000][ T0] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.16.0-mainline-06981-g076c855b846e #1
[ 0.000000][ T0] Hardware name: MT6873 (DT)
[ 0.000000][ T0] Call trace:
[ 0.000000][ T0] dump_backtrace+0xfc/0x1dc
[ 0.000000][ T0] dump_stack_lvl+0xa8/0x11c
[ 0.000000][ T0] panic+0x194/0x464
[ 0.000000][ T0] __cfi_check_fail+0x54/0x58
[ 0.000000][ T0] __cfi_slowpath_diag+0x354/0x4b0
[ 0.000000][ T0] blake2s_update+0x14c/0x178
[ 0.000000][ T0] _extract_entropy+0xf4/0x29c
[ 0.000000][ T0] crng_initialize_primary+0x24/0x94
[ 0.000000][ T0] rand_initialize+0x2c/0x6c
[ 0.000000][ T0] start_kernel+0x2f8/0x65c
[ 0.000000][ T0] __primary_switched+0xc4/0x7be4
[ 0.000000][ T0] Rebooting in 5 seconds..
Nonetheless, the function pointer method isn't so terrific anyway, so
this patch replaces it with a simple boolean, which also gets inlined
away. This successfully works around the Clang bug.
In general, I'm not too keen on all of the indirection involved here; it
clearly does more harm than good. Hopefully the whole thing can get
cleaned up down the road when lib/crypto is overhauled more
comprehensively. But for now, we go with a simple bandaid.
Link: https://lore.kernel.org/all/20220124192849.14755-1-Jason@zx2c4.com/
Bug: 218328931
Fixes: 6048fdcc5f ("lib/crypto: blake2s: include as built-in")
Link: https://github.com/ClangBuiltLinux/linux/issues/1567
Reported-by: Miles Chen <miles.chen@mediatek.com>
Tested-by: Miles Chen <miles.chen@mediatek.com>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Tested-by: John Stultz <john.stultz@linaro.org>
Acked-by: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
(cherry picked from commit d2a02e3c8b)
Signed-off-by: Will McVicker <willmcvicker@google.com>
Change-Id: I797c6d79b120041aeb98ae68060d8cc4bd165c1a
This is part of a migration process from build/
to build/kernel.
Test: builds
Bug: 204425264
Change-Id: Id29f8e476fbb3590196789b2b578865798f24cd3
Signed-off-by: Yifan Hong <elsk@google.com>
This is part of the build -> build/kernel transition.
Test: TH
Bug: 204425264
Change-Id: I5bcf6ac00b6388af198e486f191e13f0b60050c9
Signed-off-by: Yifan Hong <elsk@google.com>
Reason for revert: need alternative deployment path for test modules
Change-Id: If2d5ca5a5cd41b2b76114e2db29e6e633f6c7ec2
Bug: 181024194
Signed-off-by: Steve Muckle <smuckle@google.com>
The old and new mount user name spaces need to be populated
before calling vfs_rename().
Bug: 211066171
Signed-off-by: Tadeusz Struk <tadeusz.struk@linaro.org>
Change-Id: Ieac6975abb4131c8f5bdefe25b5f241c80023e38
Cleanup incremental-fs left overs on umount, otherwise incr-fs will
complain as below:
BUG: Dentry {i=47a,n=.incomplete} still in use [unmount of incremental-fs]
This requires vfs_rmdir() of the special index and incomplete dirs.
Also free options.sysfs_name in incfs_mount_fs() instead of in
incfs_free_mount_info() to make it consistent with incfs_remount_fs().
Since set_anon_super() was used in incfs_mount_fs() the incfs_kill_sb()
should use kill_anon_super() instead of generic_shutdown_super()
otherwise it will leak the pseudo dev_t that set_anon_super() allocates.
Bug: 211066171
Signed-off-by: Tadeusz Struk <tadeusz.struk@linaro.org>
Change-Id: I7ea54db63513fc130e1997cbf79121015ee12405
Steps on the way to 5.17-rc1
Resolves conflicts in:
drivers/scsi/ufs/ufshcd.c
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I64ad2fb6ed77ba4678f631f4143b64339b1a5ddd
This reverts commit d9a2a3f2c2 which was
not upstream as it is causing merge conflicts with 5.17-rc1.
If this out-of-tree feature is still needed, it must be forward ported
properly again.
Bug: 120440972
Cc: Doug Anderson <dianders@chromium.org>
Cc: Colin Cross <ccross@android.com>
Cc: Amit Pundir <amit.pundir@linaro.org>
Fixes: d9a2a3f2c2 ("ANDROID: of: Support CONFIG_CMDLINE_EXTEND config option")
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: Id34ae0303eb943668418b9aab3f928a7d970ca97
Steps on the way to 5.17-rc1
Resolves conflicts in:
drivers/iommu/dma-iommu.c
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I1f4b257550955dc7a79cb8a6628cd308fe2d8e71
Steps on the way to 5.17-rc1
Resolves conflicts in:
fs/ext4/inode.c
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I7f7d4309bbce5cea381efeee8c27a3505e74f0df
git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse") into
android-mainline
Steps on the way to 5.17-rc1
Signed-off-by: Alessio Balsini <balsini@google.com>
Change-Id: Iaa0fdd23d98a0a83e842dca032a14df445f87661
Starting with FUSE 7.36, all the fields for the 32-bit FUSE init flags
have been allocated, so commit 53db28933e ("fuse: extend init flags")
introduces the new 32-bit flags2 field in fuse_init_in and
fuse_init_out. That change also adds the FUSE_INIT_RESERVED flag that
doesn't have any specific purpose yet, is just reserved and should not
be used, and (un)fortunately collides with FUSE_PASSTHROUGH.
This change fixes the conflict by simply setting the FUSE_PASSTHROUGH
value to the next, latest unused fuse2 bit.
Although this is not the best design choice, userspace will know what
FUSE_PASSTHROUGH bit to choose based on the FUSE major and minor version
for FUSE version:
- < 7.36: FUSE_PASSTHROUGH is the 31st bit of flags;
- otherwise: FUSE_PASSTHROUGH is the 31st bit of flags2.
Test: launch_cvd (both android-mainline and android13-5.10) \
`logcat FuseDaemon:V \*:S` shows no FUSE passthrough errors
Bug: 215310351
Signed-off-by: Alessio Balsini <balsini@google.com>
Change-Id: I85d7582008b8c093b3172b3f41c6cdf09863dd45
Add support to install the modules.order file for external modules
during module_install in order to retain the Makefile ordering
of external modules. This helps reduce the extra steps necessary to
properly order loading of external modules when there are multiple
kernel modules compiled within a given KBUILD_EXTMOD directory.
To handle compiling multiple external modules within the same
INSTALL_MOD_DIR, kbuild will append a suffix to the installed
modules.order file defined like so:
echo "${KBUILD_EXTMOD}" | md5sum | cut -d " " -f 1
Example:
KBUILD_EXTMOD=/mnt/a.b/c-d/my_driver results in:
modules.order.7dd3eb90588c21ac15f23a96c2f6d8ec
The installed module.order.$(extmod_suffix) files can then be appended
to the staging modules.order file which defines the order to load all of
the modules during boot.
Example:
cd $(MODLIB)
find extra/. -name modules.order.* -exec cat {} >> modules.order \;
Link: https://lore.kernel.org/all/20220127010009.2617569-1-willmcvicker@google.com/
Bug: 216462633
Bug: 210713925
Signed-off-by: Will McVicker <willmcvicker@google.com>
Change-Id: I7baa92163f8e6ea3f47d780b728167d86cc2f6e1
This reverts commit 77cbbcd16f. This patch
hasn't landed upstream yet, but needs a fix. So reverting it first to
help keep the fixes together in case we need to forward port this
feature.
Bug: 216462633
Change-Id: Ia96084bd24d0ee27b7addc2acbd5277e46cb86f1
Signed-off-by: Will McVicker <willmcvicker@google.com>
This reverts commit bf2290a48a (Revert "ANDROID: vendor_hooks: set
debugging data when rt_mutex is working")
The original patch has been reverted to resolve merge issues.
This patch adds again the vendor hooks for the original purpose.
Bug: 216016261
Signed-off-by: Sangmoon Kim <sangmoon.kim@samsung.com>
Change-Id: I00162d88e2a446e9ece4804def098fcdc63fceb9
This reverts commit 31c9ccb138 (Revert "ANDROID: vendor_hooks: add
waiting information for blocked tasks")
And also revert portions of 396a501b17 (Revert "ANDROID: rwsem: Add
vendor hook to the rw-semaphore")
The original patch has been reverted to resolve merge issues.
This patch adds again the vendor hooks for the original purpose.
Bug: 216016261
Signed-off-by: Sangmoon Kim <sangmoon.kim@samsung.com>
Change-Id: I04ed7b055eee40f7975bd5d74fb73dd080cd76bf
