irq-armada-370-xp driver already sets MSI_FLAG_MULTI_PCI_MSI flag into
msi_domain_info structure. But allocated interrupt numbers for Multi-MSI
needs to be properly aligned otherwise devices send MSI interrupt with
wrong number.
Fix this issue by using function bitmap_find_free_region() instead of
bitmap_find_next_zero_area() to allocate aligned interrupt numbers.
Signed-off-by: Pali Rohár <pali@kernel.org>
Fixes: a71b9412c9 ("irqchip/armada-370-xp: Allow allocation of multiple MSIs")
Cc: stable@vger.kernel.org
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20211125130057.26705-2-pali@kernel.org
The build warning:
block/blk-core.c:968: warning: Function parameter or member 'iob'
not described in 'bio_poll'.
Fixes: 5a72e899ce ("block: add a struct io_comp_batch argument to fops->iopoll()")
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: Yang Guang <yang.guang5@zte.com.cn>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It hangup when booting Loongson 3A1000 with BOTH
CONFIG_PAGE_SIZE_64KB and CONFIG_MIPS_VA_BITS_48, that it turn
out to use 2-level pgtable instead of 3-level. 64KB page size
with 2-level pgtable only cover 42 bits VA, use 3-level pgtable
to cover all 48 bits VA(55 bits)
Fixes: 1e321fa917 ("MIPS64: Support of at least 48 bits of SEGBITS)
Signed-off-by: Huang Pei <huangpei@loongson.cn>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Having a signed (1 << 31) constant for TCR_EL2_RES1 and CPTR_EL2_TCPAC
causes the upper 32-bit to be set to 1 when assigning them to a 64-bit
variable. Bit 32 in TCR_EL2 is no longer RES0 in ARMv8.7: with FEAT_LPA2
it changes the meaning of bits 49:48 and 9:8 in the stage 1 EL2 page
table entries. As a result of the sign-extension, a non-VHE kernel can
no longer boot on a model with ARMv8.7 enabled.
CPTR_EL2 still has the top 32 bits RES0 but we should preempt any future
problems
Make these top bit constants unsigned as per commit df655b75c4
("arm64: KVM: Avoid setting the upper 32 bits of VTCR_EL2 to 1").
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Reported-by: Chris January <Chris.January@arm.com>
Cc: <stable@vger.kernel.org>
Cc: Will Deacon <will@kernel.org>
Cc: Marc Zyngier <maz@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20211125152014.2806582-1-catalin.marinas@arm.com
It turns out that 'decode_configs' -> 'set_ftlb_enable' is called under
c->cputype unset, which leaves FTLB disabled on BOTH 3A2000 and 3A3000
Fix it by calling "decode_configs" after c->cputype is initialized
Fixes: da1bd29742 ("MIPS: Loongson64: Probe CPU features via CPUCFG")
Signed-off-by: Huang Pei <huangpei@loongson.cn>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
There exists the following issue under DEBUG_PREEMPT:
BUG: using smp_processor_id() in preemptible [00000000] code: systemd/1
caller is show_cpuinfo+0x460/0xea0
...
Call Trace:
[<ffffffff8020f0dc>] show_stack+0x94/0x128
[<ffffffff80e6cab4>] dump_stack_lvl+0x94/0xd8
[<ffffffff80e74c5c>] check_preemption_disabled+0x104/0x110
[<ffffffff802209c8>] show_cpuinfo+0x460/0xea0
[<ffffffff80539d54>] seq_read_iter+0xfc/0x4f8
[<ffffffff804fcc10>] new_sync_read+0x110/0x1b8
[<ffffffff804ff57c>] vfs_read+0x1b4/0x1d0
[<ffffffff804ffb18>] ksys_read+0xd0/0x110
[<ffffffff8021c090>] syscall_common+0x34/0x58
We can see the following call trace:
show_cpuinfo()
cpu_has_fpu
current_cpu_data
smp_processor_id()
$ addr2line -f -e vmlinux 0xffffffff802209c8
show_cpuinfo
arch/mips/kernel/proc.c:188
$ head -188 arch/mips/kernel/proc.c | tail -1
if (cpu_has_fpu)
arch/mips/include/asm/cpu-features.h
# define cpu_has_fpu (current_cpu_data.options & MIPS_CPU_FPU)
arch/mips/include/asm/cpu-info.h
#define current_cpu_data cpu_data[smp_processor_id()]
Based on the above analysis, fix the issue by using raw_cpu_has_fpu
which calls raw_smp_processor_id() in show_cpuinfo().
Fixes: 626bfa0372 ("MIPS: kernel: proc: add CPU option reporting")
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Oleksandr brought a bug report where netpoll causes trace
messages in the log on igb.
Danielle brought this back up as still occurring, so we'll try
again.
[22038.710800] ------------[ cut here ]------------
[22038.710801] igb_poll+0x0/0x1440 [igb] exceeded budget in poll
[22038.710802] WARNING: CPU: 12 PID: 40362 at net/core/netpoll.c:155 netpoll_poll_dev+0x18a/0x1a0
As Alex suggested, change the driver to return work_done at the
exit of napi_poll, which should be safe to do in this driver
because it is not polling multiple queues in this single napi
context (multiple queues attached to one MSI-X vector). Several
other drivers contain the same simple sequence, so I hope
this will not create new problems.
Fixes: 16eb8815c2 ("igb: Refactor clean_rx_irq to reduce overhead and improve performance")
Reported-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Reported-by: Danielle Ratson <danieller@nvidia.com>
Suggested-by: Alexander Duyck <alexander.duyck@gmail.com>
Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Tested-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Tested-by: Danielle Ratson <danieller@nvidia.com>
Link: https://lore.kernel.org/r/20211123204000.1597971-1-jesse.brandeburg@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit c99ca78d67 ("platform/x86: thinkpad_acpi: Switch to common use
of attributes") removed the conditional adding of the
hotkey_tablet_mode and hotkey_radio_sw sysfs-attributes, replacing this
with a hotkey_attr_is_visible() callback which hides them when the
feature is not present.
But this commit forgot to add these 2 attributes to the default
hotkey_attributes[] set, so they would now never get added at all.
Add the 2 attributes to the default hotkey_attributes[] set so that
they are available on systems with these features once more.
Fixes: c99ca78d67 ("platform/x86: thinkpad_acpi: Switch to common use of attributes")
Cc: Len Baker <len.baker@gmx.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20211123210424.266607-2-hdegoede@redhat.com
Pull NVMe fixes from Christoph:
"nvme fixes for Linux 5.16
- add a NO APST quirk for a Kioxia device (Enzo Matsumiya)
- fix write zeroes pi (Klaus Jensen)
- various TCP transport fixes (Maurizio Lombardi and Varun Prakash)
- ignore invalid fast_io_fail_tmo values (Maurizio Lombardi)
- use IOCB_NOWAIT only if the filesystem supports it (Maurizio Lombardi)"
* tag 'nvme-5.16-2021-11-25' of git://git.infradead.org/nvme:
nvmet: use IOCB_NOWAIT only if the filesystem supports it
nvme: fix write zeroes pi
nvme-fabrics: ignore invalid fast_io_fail_tmo values
nvme-pci: add NO APST quirk for Kioxia device
nvme-tcp: fix memory leak when freeing a queue
nvme-tcp: validate R2T PDU in nvme_tcp_handle_r2t()
nvmet-tcp: fix incomplete data digest send
nvmet-tcp: fix memory leak when performing a controller reset
nvmet-tcp: add an helper to free the cmd buffers
nvmet-tcp: fix a race condition between release_queue and io_work
We're hitting OOB accesses in handle_ipmb_direct_rcv_rsp() (memcpy of
size -1) after user space generates a message. Looks like the message
is incorrectly assumed to be of the new IPMB type, because type is never
set and message is allocated with kmalloc() not kzalloc().
Fixes: 059747c245 ("ipmi: Add support for IPMB direct messages")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Message-Id: <20211124210323.1950976-1-kuba@kernel.org>
Signed-off-by: Corey Minyard <cminyard@mvista.com>
Submit I/O requests with the IOCB_NOWAIT flag set only if
the underlying filesystem supports it.
Fixes: 50a909db36 ("nvmet: use IOCB_NOWAIT for file-ns buffered I/O")
Signed-off-by: Maurizio Lombardi <mlombard@redhat.com>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
We add the ostd setting for mt8195. It introduces a KE for the
previous SoC which doesn't have ostd setting. This is the log:
Unable to handle kernel NULL pointer dereference at virtual address
0000000000000080
...
pc : mtk_smi_larb_config_port_gen2_general+0x64/0x130
lr : mtk_smi_larb_resume+0x54/0x98
...
Call trace:
mtk_smi_larb_config_port_gen2_general+0x64/0x130
pm_generic_runtime_resume+0x2c/0x48
__genpd_runtime_resume+0x30/0xa8
genpd_runtime_resume+0x94/0x2c8
__rpm_callback+0x44/0x150
rpm_callback+0x6c/0x78
rpm_resume+0x310/0x558
__pm_runtime_resume+0x3c/0x88
In the code: larbostd = larb->larb_gen->ostd[larb->larbid],
if "larb->larb_gen->ostd" is null, the "larbostd" is the offset(e.g.
0x80 above), it's also a valid value, then accessing "larbostd[i]" in the
"for" loop will cause the KE above. To avoid this issue, initialize
"larbostd" to NULL when the SoC doesn't have ostd setting.
Fixes: fe6dd2a401 ("memory: mtk-smi: mt8195: Add initial setting for smi-larb")
Signed-off-by: Yong Wu <yong.wu@mediatek.com>
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20211108082429.15080-1-yong.wu@mediatek.com
Link: https://lore.kernel.org/r/20211124085042.9649-3-krzysztof.kozlowski@canonical.com'
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
ASoC: Fixes for v5.16
There's a large but repetitive set of fixes here for issues with the
Tegra kcontrols not correctly reporting changes to userspace, a fix for
some issues with matching on older x86 platforms introduced during the
merge window together with a set of smaller fixes and one new system
quirk.
The signaled bit is already used for quick testing if a fence is signaled.
On top of that, it's a terrible abuse of dma-fence api, and in the common
case where the object is already locked by the caller, the trylock will fail.
If it were useful, the core dma-api would have exposed the same functionality.
The fact that i915 has a dma_resv_utils.c file should be a warning that the
functionality either belongs in core, or is not very useful at all.
In this case the latter.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
[mlankhorst: Improve commit message]
Acked-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211021103605.735002-3-maarten.lankhorst@linux.intel.com
Reviewed-by: Matthew Auld <matthew.auld@intel.com> #irc
Checking buf->flags should be done before the pipe_buf_release() is called
on the pipe buffer, since releasing the buffer might modify the flags.
This is exactly what page_cache_pipe_buf_release() does, and which results
in the same VM_BUG_ON_PAGE(PageLRU(page)) that the original patch was
trying to fix.
Reported-by: Justin Forbes <jmforbes@linuxtx.org>
Fixes: 712a951025 ("fuse: fix page stealing")
Cc: <stable@vger.kernel.org> # v2.6.35
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>
Maarten requested a backmerge due his work depending on subtle semantic
changes introduced by:
7e2e69ed46 ("drm/i915: Fix i915_request fence wait semantics")
2cbb8d4d67 ("drm/i915: use new iterator in i915_gem_object_wait_reservation")
Both should probably have been merged to drm-intel-gt-next anyway.
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Johan writes:
USB-serial fixes for 5.16-rc3
Here's a fix for a pl2303 type-detection issue and some new modem
devices ids.
All but the last commit have been in linux-next, and with no reported
issues.
* tag 'usb-serial-5.16-rc3' of https://git.kernel.org/pub/scm/linux/kernel/git/johan/usb-serial:
USB: serial: option: add Fibocom FM101-GL variants
USB: serial: pl2303: fix GC type detection
USB: serial: option: add Telit LE910S1 0x9200 composition
Both, hardware and drivers does support interrupts.
Fix warnings as:
arch/arm/boot/dts/tegra30-microsoft-surface-rt-efi.dt.yaml: audio-codec@1a: 'interrupt-parent', 'interrupts' do not match any of the regexes: 'pinctrl-[0-9]+'
From schema: /home/runner/work/linux/linux/Documentation/devicetree/bindings/sound/wlf,wm8962.yaml
Fixes: cd51b942f3 ("ASoC: dt-bindings: wlf,wm8962: Convert to json-schema")
Signed-off-by: David Heidelberg <david@ixit.cz>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/r/20211124155101.59694-1-david@ixit.cz
Signed-off-by: Mark Brown <broonie@kernel.org>
Just like before with __bswapdi2(), for MIPS pre-boot when
CONFIG_KERNEL_ZSTD=y the decompressor function will use __ashldi3(), so
the object file should be added to the target object file.
Fixes these build errors:
mipsel-linux-ld: arch/mips/boot/compressed/decompress.o: in function `FSE_buildDTable_internal':
decompress.c:(.text.FSE_buildDTable_internal+0x48): undefined reference to `__ashldi3'
mipsel-linux-ld: arch/mips/boot/compressed/decompress.o: in function `FSE_decompress_wksp_body_default':
decompress.c:(.text.FSE_decompress_wksp_body_default+0xa8): undefined reference to `__ashldi3'
mipsel-linux-ld: arch/mips/boot/compressed/decompress.o: in function `ZSTD_getFrameHeader_advanced':
decompress.c:(.text.ZSTD_getFrameHeader_advanced+0x134): undefined reference to `__ashldi3'
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Don't wait sync while migrating, but rather make the GPU blit await the
dependencies and add a moving fence to the object.
This also enables asynchronous VRAM management in that on eviction,
rather than waiting for the moving fence to expire before freeing VRAM,
it is freed immediately and the fence is stored with the VRAM manager and
handed out to newly allocated objects to await before clears and swapins,
or for kernel objects before setting up gpu vmas or mapping.
To collect dependencies before migrating, add a set of utilities that
coalesce these to a single dma_fence.
What is still missing for fully asynchronous operation is asynchronous vma
unbinding, which is still to be implemented.
This commit substantially reduces execution time in the gem_lmem_swapping
test.
v2:
- Make a couple of functions static.
v4:
- Fix some style issues (Matthew Auld)
- Audit and add more checks for ghost objects (Matthew Auld)
- Add more documentation for the i915_deps utility (Mattew Auld)
- Simplify the i915_deps_sync() function
v6:
- Re-check for fence signaled before returning -EBUSY (Matthew Auld)
- Use dma_resv_iter_is_exclusive() (Matthew Auld)
- Await all dma-resv fences before a migration blit (Matthew Auld)
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211122214554.371864-6-thomas.hellstrom@linux.intel.com
With async migration, the shrinker may end up wanting to release the
pages of an object while the migration blit is still running, since
the GT migration code doesn't set up VMAs and the shrinker is thus
oblivious to the fact that the GPU is still using the pages.
Add waiting for gpu in the shrinker_release_pages() op and an
argument to that function indicating whether the shrinker expects it
to not wait for gpu. In the latter case the shrinker_release_pages()
op will return -EBUSY if the object is not idle.
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211122214554.371864-5-thomas.hellstrom@linux.intel.com
There is an interesting refcounting loop:
struct intel_memory_region has a struct ttm_resource_manager,
ttm_resource_manager->move may hold a reference to i915_request,
i915_request may hold a reference to intel_context,
intel_context may hold a reference to drm_i915_gem_object,
drm_i915_gem_object may hold a reference to intel_memory_region.
Break this loop by dropping region reference counting.
In addition, Have regions with a manager moving fence make sure
that all region objects are released before freeing the region.
v6:
- Fix a code comment.
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211122214554.371864-4-thomas.hellstrom@linux.intel.com
For now, we will only allow async migration when TTM is used,
so the paths we care about are related to TTM.
The mmap path is handled by having the fence in ttm_bo->moving,
when pinning, the binding only becomes available after the moving
fence is signaled, and pinning a cpu map will only work after
the moving fence signals.
This should close all holes where userspace can read a buffer
before it's fully migrated.
v2:
- Fix a couple of SPARSE warnings
v3:
- Fix a NULL pointer dereference
v4:
- Ditch the moving fence waiting for i915_vma_pin_iomap() and
replace with a verification that the vma is already bound.
(Matthew Auld)
- Squash with a previous patch introducing moving fence waiting and
accessing interfaces (Matthew Auld)
- Rename to indicated that we also add support for sync waiting.
v5:
- Fix check for NULL and unreferencing i915_vma_verify_bind_complete()
(Matthew Auld)
- Fix compilation failure if !CONFIG_DRM_I915_DEBUG_GEM
- Fix include ordering. (Matthew Auld)
v7:
- Fix yet another compilation failure with clang if
!CONFIG_DRM_I915_DEBUG_GEM
Co-developed-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20211122214554.371864-2-thomas.hellstrom@linux.intel.com
If xattr is not supported like exfat or fat, ksmbd server doesn't
contain default data stream in FILE_STREAM_INFORMATION response. It will
cause ppt or doc file update issue if local filesystem is such as ones.
This patch move goto statement to contain it.
Fixes: 9f6323311c ("ksmbd: add default data stream name in FILE_STREAM_INFORMATION")
Cc: stable@vger.kernel.org # v5.15
Acked-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
While file transfer through windows client, This error flood message
happen. This flood message will cause performance degradation and
misunderstand server has problem.
Fixes: e294f78d34 ("ksmbd: allow PROTECTED_DACL_SECINFO and UNPROTECTED_DACL_SECINFO addition information in smb2 set info security")
Cc: stable@vger.kernel.org # v5.15
Acked-by: Hyunchul Lee <hyc.lee@gmail.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
All the error handling paths of 'smb2_sess_setup()' end to 'out_err'.
All but the new error handling path added by the commit given in the Fixes
tag below.
Fix this error handling path and branch to 'out_err' as well.
Fixes: 0d994cd482 ("ksmbd: add buffer validation in session setup")
Cc: stable@vger.kernel.org # v5.15
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Steve French <stfrench@microsoft.com>
Karsten Graul says:
====================
net/smc: fixes 2021-11-24
Patch 1 from DaXing fixes a possible loop in smc_listen().
Patch 2 prevents a NULL pointer dereferencing while iterating
over the lower network devices.
====================
Link: https://lore.kernel.org/r/20211124123238.471429-1-kgraul@linux.ibm.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The kernel_listen function in smc_listen will fail when all the available
ports are occupied. At this point smc->clcsock->sk->sk_data_ready has
been changed to smc_clcsock_data_ready. When we call smc_listen again,
now both smc->clcsock->sk->sk_data_ready and smc->clcsk_data_ready point
to the smc_clcsock_data_ready function.
The smc_clcsock_data_ready() function calls lsmc->clcsk_data_ready which
now points to itself resulting in an infinite loop.
This patch restores smc->clcsock->sk->sk_data_ready with the old value.
Fixes: a60a2b1e0a ("net/smc: reduce active tcp_listen workers")
Signed-off-by: Guo DaXing <guodaxing@huawei.com>
Acked-by: Tony Lu <tonylu@linux.alibaba.com>
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Coverity reports a possible NULL dereferencing problem:
in smc_vlan_by_tcpsk():
6. returned_null: netdev_lower_get_next returns NULL (checked 29 out of 30 times).
7. var_assigned: Assigning: ndev = NULL return value from netdev_lower_get_next.
1623 ndev = (struct net_device *)netdev_lower_get_next(ndev, &lower);
CID 1468509 (#1 of 1): Dereference null return value (NULL_RETURNS)
8. dereference: Dereferencing a pointer that might be NULL ndev when calling is_vlan_dev.
1624 if (is_vlan_dev(ndev)) {
Remove the manual implementation and use netdev_walk_all_lower_dev() to
iterate over the lower devices. While on it remove an obsolete function
parameter comment.
Fixes: cb9d43f677 ("net/smc: determine vlan_id of stacked net_device")
Suggested-by: Julian Wiedmann <jwi@linux.ibm.com>
Signed-off-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Marek Behún says:
====================
phylink resolve fixes
With information from me and my nagging, Russell has produced two fixes
for phylink, which add code that triggers another phylink_resolve() from
phylink_resolve(), if certain conditions are met:
interface is being changed
or
link is down and previous link was up
These are needed because sometimes the PCS callbacks may provide stale
values if link / speed / ...
====================
Link: https://lore.kernel.org/r/20211123154403.32051-1-kabel@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
On mv88e6xxx 1G/2.5G PCS, the SerDes register 4.2001.2 has the following
description:
This register bit indicates when link was lost since the last
read. For the current link status, read this register
back-to-back.
Thus to get current link state, we need to read the register twice.
But doing that in the link change interrupt handler would lead to
potentially ignoring link down events, which we really want to avoid.
Thus this needs to be solved in phylink's resolve, by retriggering
another resolve in the event when PCS reports link down and previous
link was up, and by re-reading PCS state if the previous link was down.
The wrong value is read when phylink requests change from sgmii to
2500base-x mode, and link won't come up. This fixes the bug.
Fixes: 9525ae8395 ("phylink: add phylink infrastructure")
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Marek Behún <kabel@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
On PHY state change the phylink_resolve() function can read stale
information from the MAC and report incorrect link speed and duplex to
the kernel message log.
Example with a Marvell 88X3310 PHY connected to a SerDes port on Marvell
88E6393X switch:
- PHY driver triggers state change due to PHY interface mode being
changed from 10gbase-r to 2500base-x due to copper change in speed
from 10Gbps to 2.5Gbps, but the PHY itself either hasn't yet changed
its interface to the host, or the interrupt about loss of SerDes link
hadn't arrived yet (there can be a delay of several milliseconds for
this), so we still think that the 10gbase-r mode is up
- phylink_resolve()
- phylink_mac_pcs_get_state()
- this fills in speed=10g link=up
- interface mode is updated to 2500base-x but speed is left at 10Gbps
- phylink_major_config()
- interface is changed to 2500base-x
- phylink_link_up()
- mv88e6xxx_mac_link_up()
- .port_set_speed_duplex()
- speed is set to 10Gbps
- reports "Link is Up - 10Gbps/Full" to dmesg
Afterwards when the interrupt finally arrives for mv88e6xxx, another
resolve is forced in which we get the correct speed from
phylink_mac_pcs_get_state(), but since the interface is not being
changed anymore, we don't call phylink_major_config() but only
phylink_mac_config(), which does not set speed/duplex anymore.
To fix this, we need to force the link down and trigger another resolve
on PHY interface change event.
Fixes: 9525ae8395 ("phylink: add phylink infrastructure")
Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Marek Behún <kabel@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
