unlike other classifiers that can be offloaded (i.e. users can set flags
like 'skip_hw' and 'skip_sw'), 'cls_matchall' doesn't validate the size
of netlink attribute 'TCA_MATCHALL_FLAGS' provided by user: add a proper
entry to mall_policy.
Fixes: b87f7936a9 ("net/sched: Add match-all classifier hw offloading.")
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
@thoff has moved to struct flow_dissector_key_control.
Fixes: 42aecaa9bb ("net: Get skb hash over flow_keys structure")
Signed-off-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
"do {} while" in page_pool_refill_alloc_cache will always
refill page once whether refill is true or false, and whether
alloc.count of pool is less than PP_ALLOC_CACHE_REFILL or not
this is wrong, and will cause overflow of pool->alloc.cache
the caller of __page_pool_get_cached should provide guarantee
that pool->alloc.cache is safe to access, so in_serving_softirq
should be removed as suggested by Jesper Dangaard Brouer in
https://patchwork.ozlabs.org/patch/1233713/
so fix this issue by calling page_pool_refill_alloc_cache()
only when pool->alloc.count is zero
Fixes: 44768decb7 ("page_pool: handle page recycle for NUMA_NO_NODE condition")
Signed-off-by: Li RongQing <lirongqing@baidu.com>
Suggested-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jeff Kirsher says:
====================
Intel Wired LAN Driver Updates 2020-02-12
This series contains fixes to only the ice driver.
Dave fixes logic flaws in the DCB rebuild function which is used after a
reset. Also fixed a configuration issue when switching between firmware
and software LLDP mode where the number of TLV's configured was getting
out of sync with what lldpad thinks is configured.
Paul fixes how the driver displayed all the supported and advertised
link modes by basing it on the PHY capabilities, and in the process
cleaned up a lot of code.
Brett fixes duplicate receive tail bumps by comparing the value we are
writing to tail with the previously written tail value. Also cleaned up
workarounds that are no longer needed with the latest NVM images.
Anirudh cleaned up unnecessary CONFIG_PCI_IOV wrappers. Updated the
driver to use ice_pf_to_dev() instead of &pf->pdev->dev or
&vsi->back->pdev->dev. Cleaned up the string format in print function
calls to remove newlines where applicable.
Akeem updates the link message logging to include "Full Duplex" and
"Negotiated", to help distinguish from "Requested" for FEC.
Bruce fixes and consolidates the logging of firmware/NVM information
during driver load, since the information is duplicate of what is
available via ethtool. Fixed the checking of the Unit Load Status bits
after reset to ensure they are 0x7FF before continuing, by updating the
mask. Cleanup up possible NULL dereferences that were created by a
previous commit.
Ben fixes the driver to use the correct netif_msg_tx/rx_error() to
determine whether to print the MDD event type.
Tony provides several trivial fixes, which include whitespace, typos,
function header comments, reverse Christmas tree issues.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Move vga switcheroo and dsm handler register later in
i915_driver_register(), and unregister in i915_driver_unregister(). The
dsm handler unregister is a nop, and is only added for completeness.
My unsubstantiated suspicion is that the vga switcheroo state change
would not work as early as we register the hooks currently. In any case
exposing the interfaces to the world only after we've got everything set
up seems prudent.
Also replace the error handling in vga switcheroo register with a simple
error message. This is done at the same time due to lack of error
propagation from i915_driver_register().
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Acked-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211162802.16180-1-jani.nikula@intel.com
The component order between the two was swapped, resulting in incorrect
color when games with 565 visual hit the overlay path instead of GPU
composition.
Fixes: 25fdd5933e ("drm/msm: Add SDM845 DPU support")
Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
For a given byte clock, if VCO recalc value is exactly same as
vco set rate value, vco_set_rate does not get called assuming
VCO is already set to required value. But Due to GDSC toggle,
VCO values are erased in the HW. To make sure VCO is programmed
correctly, we forcefully call set_rate from vco_prepare.
Signed-off-by: Harigovindan P <harigovi@codeaurora.org>
Reviewed-by: Jeffrey Hugo <jeffrey.l.hugo@gmail.com>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Save pll state before dsi host is powered off. Without this change
some register values gets resetted.
Signed-off-by: Harigovindan P <harigovi@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Ensure we don't release the delegation cred during the call to
nfs4_proc_delegreturn().
Fixes: ee05f45677 ("NFSv4: Fix races between open and delegreturn")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The call to nfs_do_return_delegation() needs to be taken without
any RCU locks. Add a refcount to make sure the delegation remains
pinned in memory until we're done.
Fixes: ee05f45677 ("NFSv4: Fix races between open and delegreturn")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Since 'smu_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)' will always return
false considering the 'smu_system_features_control(smu, false)' disabled
all SMU features.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Kenneth Feng <kenneth.feng@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This prepares the addition of the GX SoC family sound card driver.
The GX sound card, while slightly different, will be similar to the
AXG one. The purpose of this change is to share the utils common to
both sound card driver.
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20200213155159.3235792-8-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add the codec to codec component which handles the routing between
the audio producers and the internal audio DAC found on the amlogic GXL
SoC family
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20200213155159.3235792-7-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add the codec to codec component which handles the routing between
the audio producers (PCM and I2S) and the synopsys hdmi controller
on the amlogic GX SoC family
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20200213155159.3235792-6-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Add the dt-bindings and documentation of the AIU audio controller.
This component provides most of the audio outputs found on the Amlogic
Gx SoC family.
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20200213155159.3235792-4-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The hdmi routing mechanism used on g12a hdmi is also used:
* other Amlogic SoC types
* for the internal DAC path
Each of these codec glues are slightly different but the idea
behind it remains the same. This change extract some helper functions
from the g12a-tohdmitx driver to make them available for other Amlogic
codecs.
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20200213155159.3235792-3-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
At the moment, querying the dai_name will stop of the first component
matching the dt node. This does not allow a device (single dt node) to
provide several ASoC components which could then be used through DT.
This change let the search go on if the xlate function of the component
returns an error, giving the possibility to another component to match
and return the dai_name.
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20200213155159.3235792-2-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
HiP08 RoCE hardware lacks ability(a known hardware problem) to flush
outstanding WQEs if QP state gets into errored mode for some reason. To
overcome this hardware problem and as a workaround, when QP is detected to
be in errored state during various legs like post send, post receive
etc[1], flush needs to be performed from the driver.
The earlier patch[1] sent to solve the hardware limitation explained in
the cover-letter had a bug in the software flushing leg. It acquired mutex
while modifying QP state to errored state and while conveying it to the
hardware using the mailbox. This caused leg to sleep while holding
spin-lock and caused crash.
Suggested Solution: we have proposed to defer the flushing of the QP in
the Errored state using the workqueue to get around with the limitation of
our hardware.
This patch specifically adds the calls to the flush handler from where
parts of the code like post_send/post_recv etc. when the QP state gets
into the errored mode.
[1] https://patchwork.kernel.org/patch/10534271/
Link: https://lore.kernel.org/r/1580983005-13899-3-git-send-email-liuyixian@huawei.com
Signed-off-by: Yixian Liu <liuyixian@huawei.com>
Reviewed-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
HiP08 RoCE hardware lacks ability(a known hardware problem) to flush
outstanding WQEs if QP state gets into errored mode for some reason. To
overcome this hardware problem and as a workaround, when QP is detected to
be in errored state during various legs like post send, post receive etc
[1], flush needs to be performed from the driver.
The earlier patch[1] sent to solve the hardware limitation explained in
the cover-letter had a bug in the software flushing leg. It acquired mutex
while modifying QP state to errored state and while conveying it to the
hardware using the mailbox. This caused leg to sleep while holding
spin-lock and caused crash.
Suggested Solution:
we have proposed to defer the flushing of the QP in the Errored state
using the workqueue to get around with the limitation of our hardware.
This patch adds the framework of the workqueue and the flush handler
function.
[1] https://patchwork.kernel.org/patch/10534271/
Link: https://lore.kernel.org/r/1580983005-13899-2-git-send-email-liuyixian@huawei.com
Signed-off-by: Yixian Liu <liuyixian@huawei.com>
Reviewed-by: Salil Mehta <salil.mehta@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Turning caching off for writes on the server should improve performance.
Fixes: fba83f3411 ("NFS: Pass "privileged" value to nfs4_init_sequence()")
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Reviewed-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The @nents value that was passed to ib_dma_map_sg() has to be passed
to the matching ib_dma_unmap_sg() call. If ib_dma_map_sg() choses to
concatenate sg entries, it will return a different nents value than
it was passed.
The bug was exposed by recent changes to the AMD IOMMU driver, which
enabled sg entry concatenation.
Looking all the way back to commit 4143f34e01 ("xprtrdma: Port to
new memory registration API") and reviewing other kernel ULPs, it's
not clear that the frwr_map() logic was ever correct for this case.
Reported-by: Andre Tomt <andre@tomt.net>
Suggested-by: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Cc: stable@vger.kernel.org
Reviewed-by: Jason Gunthorpe <jgg@mellanox.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
It appears that newer glibcs check that openat(O_CREAT) was provided a
fourth argument (rather than passing garbage), resulting in the
following build error:
> In file included from /usr/include/fcntl.h:301,
> from helpers.c:9:
> In function 'openat',
> inlined from 'touchat' at helpers.c:49:11:
> /usr/include/x86_64-linux-gnu/bits/fcntl2.h:126:4: error: call to
> '__openat_missing_mode' declared with attribute error: openat with O_CREAT
> or O_TMPFILE in third argument needs 4 arguments
> 126 | __openat_missing_mode ();
> | ^~~~~~~~~~~~~~~~~~~~~~~~
Reported-by: Shuah Khan <shuah@kernel.org>
Signed-off-by: Aleksa Sarai <cyphar@cyphar.com>
Tested-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
While building selftests, the following errors were observed:
> tools/testing/selftests/timens'
> gcc -Wall -Werror -pthread -lrt -ldl timens.c -o tools/testing/selftests/timens/timens
> /usr/bin/ld: /tmp/ccGy5CST.o: in function `check_config_posix_timers':
> timens.c:(.text+0x65a): undefined reference to `timer_create'
> collect2: error: ld returned 1 exit status
Quoting commit 870f193d48 ("selftests: net: use LDLIBS instead of
LDFLAGS"):
The default Makefile rule looks like:
$(CC) $(CFLAGS) $(LDFLAGS) $@ $^ $(LDLIBS)
When linking is done by gcc itself, no issue, but when it needs to be passed
to proper ld, only LDLIBS follows and then ld cannot know what libs to link
with.
More detail:
https://www.gnu.org/software/make/manual/html_node/Implicit-Variables.html
LDFLAGS
Extra flags to give to compilers when they are supposed to invoke the linker,
‘ld’, such as -L. Libraries (-lfoo) should be added to the LDLIBS variable
instead.
LDLIBS
Library flags or names given to compilers when they are supposed to invoke the
linker, ‘ld’. LOADLIBES is a deprecated (but still supported) alternative to
LDLIBS. Non-library linker flags, such as -L, should go in the LDFLAGS
variable.
While at here, correct other selftests, not only timens ones.
Reported-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Dmitry Safonov <dima@arista.com>
Tested-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
syzkaller reported this UAF:
BUG: KASAN: use-after-free in n_tty_receive_buf_common+0x2481/0x2940 drivers/tty/n_tty.c:1741
Read of size 1 at addr ffff8880089e40e9 by task syz-executor.1/13184
CPU: 0 PID: 13184 Comm: syz-executor.1 Not tainted 5.4.7 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
Call Trace:
...
kasan_report+0xe/0x20 mm/kasan/common.c:634
n_tty_receive_buf_common+0x2481/0x2940 drivers/tty/n_tty.c:1741
tty_ldisc_receive_buf+0xac/0x190 drivers/tty/tty_buffer.c:461
paste_selection+0x297/0x400 drivers/tty/vt/selection.c:372
tioclinux+0x20d/0x4e0 drivers/tty/vt/vt.c:3044
vt_ioctl+0x1bcf/0x28d0 drivers/tty/vt/vt_ioctl.c:364
tty_ioctl+0x525/0x15a0 drivers/tty/tty_io.c:2657
vfs_ioctl fs/ioctl.c:47 [inline]
It is due to a race between parallel paste_selection (TIOCL_PASTESEL)
and set_selection_user (TIOCL_SETSEL) invocations. One uses sel_buffer,
while the other frees it and reallocates a new one for another
selection. Add a mutex to close this race.
The mutex takes care properly of sel_buffer and sel_buffer_lth only. The
other selection global variables (like sel_start, sel_end, and sel_cons)
are protected only in set_selection_user. The other functions need quite
some more work to close the races of the variables there. This is going
to happen later.
This likely fixes (I am unsure as there is no reproducer provided) bug
206361 too. It was marked as CVE-2020-8648.
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Reported-by: syzbot+59997e8d5cbdc486e6f6@syzkaller.appspotmail.com
References: https://bugzilla.kernel.org/show_bug.cgi?id=206361
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20200210081131.23572-2-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When pasting a selection to a vt, the task is set as INTERRUPTIBLE while
waiting for a tty to unthrottle. But signals are not handled at all.
Normally, this is not a problem as tty_ldisc_receive_buf receives all
the goods and a user has no reason to interrupt the task.
There are two scenarios where this matters:
1) when the tty is throttled and a signal is sent to the process, it
spins on a CPU until the tty is unthrottled. schedule() does not
really echedule, but returns immediately, of course.
2) when the sel_buffer becomes invalid, KASAN prevents any reads from it
and the loop simply does not proceed and spins forever (causing the
tty to throttle, but the code never sleeps, the same as above). This
sometimes happens as there is a race in the sel_buffer handling code.
So add signal handling to this ioctl (TIOCL_PASTESEL) and return -EINTR
in case a signal is pending.
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: stable <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20200210081131.23572-1-jslaby@suse.cz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Christophe reports that powerpc 8xx silently fails to 5.6-rc1. It turns
out I was wrong about nobody relying on the lazy initialization of the
cpm/qe muram in commit b6231ea2b3 (soc: fsl: qe: drop broken lazy
call of cpm_muram_init()).
Rather than reinstating the somewhat dubious lazy call (initializing a
currently held spinlock, and implicitly doing a GFP_KERNEL under that
spinlock), make sure that cpm_muram_init() is called early enough - I
thought the calls from the subsys_initcalls were good enough, but when
used by console drivers, that's obviously not the
case. cpm_muram_init() is safe to call twice (there's an early return
if it is already initialized), so keep the call from cpm_init() - in
case SERIAL_CPM_CONSOLE=n.
Fixes: b6231ea2b3 (soc: fsl: qe: drop broken lazy call of cpm_muram_init())
Reported-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Tested-by: Christophe Leroy <christophe.leroy@c-s.fr>
Link: https://lore.kernel.org/r/20200213114342.21712-1-linux@rasmusvillemoes.dk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Link: https://lore.kernel.org/r/20200213004426.GA7886@embeddedor.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Link: https://lore.kernel.org/r/20200213004611.GA8748@embeddedor.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If kref_get_unless_zero() fails, we should keep looking for the
next service, since the callers of this function expect that a NULL
return value means there are no more.
Signed-off-by: Marcelo Diop-Gonzalez <marcgonzalez@google.com>
Link: https://lore.kernel.org/r/20200213194001.130110-1-marcgonzalez@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is somehow no description of %c corename format specifier for
/proc/sys/kernel/core_pattern. The %c corename format specifier is
used by user-space application such as systemd-coredump, so it should
be documented.
To find where %c is handled in the kernel source code, look at
function format_corename() in fs/coredump.c.
Signed-off-by: HATAYAMA Daisuke <d.hatayama@fujitsu.com>
Link: https://lore.kernel.org/r/TYAPR01MB4014714BB2ACE425BB6EC6B7951A0@TYAPR01MB4014.jpnprd01.prod.outlook.com
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
drm->dev_private is to be avoided. Use to_i915() on the struct
drm_device pointer instead. Rename the affected local dev_priv variables
to i915 while at it.
Cc: Wambui Karuga <wambui.karugax@gmail.com>
Reviewed-by: Wambui Karuga <wambui.karugax@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200211134427.31605-1-jani.nikula@intel.com
Convert rpc-cache.txt to ReST. Changes aim to improve presentation
but the content itself remains mostly the same.
Signed-off-by: Daniel W. S. Almeida <dwlsalmeida@gmail.com>
Link: https://lore.kernel.org/r/20200129044917.566906-3-dwlsalmeida@gmail.com
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
threshold_create_bank() creates a bank descriptor per MCA error
thresholding counter which can be controlled over sysfs. It publishes
the pointer to that bank in a per-CPU variable and then goes on to
create additional thresholding blocks if the bank has such.
However, that creation of additional blocks in
allocate_threshold_blocks() can fail, leading to a use-after-free
through the per-CPU pointer.
Therefore, publish that pointer only after all blocks have been setup
successfully.
Fixes: 019f34fccf ("x86, MCE, AMD: Move shared bank to node descriptor")
Reported-by: Saar Amar <Saar.Amar@microsoft.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: <stable@vger.kernel.org>
Link: http://lkml.kernel.org/r/20200128140846.phctkvx5btiexvbx@kili.mountain
For memory regions registered with IB_ACCESS_RELAXED_ORDERING will be dma
mapped with the DMA_ATTR_WEAK_ORDERING.
This will allow reads and writes to the mapping to be weakly ordered, such
change can enhance performance on some supporting architectures.
Link: https://lore.kernel.org/r/20200212073559.684139-1-leon@kernel.org
Signed-off-by: Michael Guralnik <michaelgur@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Commit 904cdbd41d ("jbd2: clear dirty flag when revoking a buffer from
an older transaction") set the BH_Freed flag when forgetting a metadata
buffer which belongs to the committing transaction, it indicate the
committing process clear dirty bits when it is done with the buffer. But
it also clear the BH_Mapped flag at the same time, which may trigger
below NULL pointer oops when block_size < PAGE_SIZE.
rmdir 1 kjournald2 mkdir 2
jbd2_journal_commit_transaction
commit transaction N
jbd2_journal_forget
set_buffer_freed(bh1)
jbd2_journal_commit_transaction
commit transaction N+1
...
clear_buffer_mapped(bh1)
ext4_getblk(bh2 ummapped)
...
grow_dev_page
init_page_buffers
bh1->b_private=NULL
bh2->b_private=NULL
jbd2_journal_put_journal_head(jh1)
__journal_remove_journal_head(hb1)
jh1 is NULL and trigger oops
*) Dir entry block bh1 and bh2 belongs to one page, and the bh2 has
already been unmapped.
For the metadata buffer we forgetting, we should always keep the mapped
flag and clear the dirty flags is enough, so this patch pick out the
these buffers and keep their BH_Mapped flag.
Link: https://lore.kernel.org/r/20200213063821.30455-3-yi.zhang@huawei.com
Fixes: 904cdbd41d ("jbd2: clear dirty flag when revoking a buffer from an older transaction")
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: zhangyi (F) <yi.zhang@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
There is no need to delay the clearing of b_modified flag to the
transaction committing time when unmapping the journalled buffer, so
just move it to the journal_unmap_buffer().
Link: https://lore.kernel.org/r/20200213063821.30455-2-yi.zhang@huawei.com
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: zhangyi (F) <yi.zhang@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
DIR_INDEX has been introduced as a compat ext4 feature. That means that
even kernels / tools that don't understand the feature may modify the
filesystem. This works because for kernels not understanding indexed dir
format, internal htree nodes appear just as empty directory entries.
Index dir aware kernels then check the htree structure is still
consistent before using the data. This all worked reasonably well until
metadata checksums were introduced. The problem is that these
effectively made DIR_INDEX only ro-compatible because internal htree
nodes store checksums in a different place than normal directory blocks.
Thus any modification ignorant to DIR_INDEX (or just clearing
EXT4_INDEX_FL from the inode) will effectively cause checksum mismatch
and trigger kernel errors. So we have to be more careful when dealing
with indexed directories on filesystems with checksumming enabled.
1) We just disallow loading any directory inodes with EXT4_INDEX_FL when
DIR_INDEX is not enabled. This is harsh but it should be very rare (it
means someone disabled DIR_INDEX on existing filesystem and didn't run
e2fsck), e2fsck can fix the problem, and we don't want to answer the
difficult question: "Should we rather corrupt the directory more or
should we ignore that DIR_INDEX feature is not set?"
2) When we find out htree structure is corrupted (but the filesystem and
the directory should in support htrees), we continue just ignoring htree
information for reading but we refuse to add new entries to the
directory to avoid corrupting it more.
Link: https://lore.kernel.org/r/20200210144316.22081-1-jack@suse.cz
Fixes: dbe8944404 ("ext4: Calculate and verify checksums for htree nodes")
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Cc: stable@kernel.org
