If we attempt a direct issue to a SCSI device, and it returns BUSY, then
we queue the request up normally. However, the SCSI layer may have
already setup SG tables etc for this particular command. If we later
merge with this request, then the old tables are no longer valid. Once
we issue the IO, we only read/write the original part of the request,
not the new state of it.
This causes data corruption, and is most often noticed with the file
system complaining about the just read data being invalid:
[ 235.934465] EXT4-fs error (device sda1): ext4_iget:4831: inode #7142: comm dpkg-query: bad extra_isize 24937 (inode size 256)
because most of it is garbage...
This doesn't happen from the normal issue path, as we will simply defer
the request to the hardware queue dispatch list if we fail. Once it's on
the dispatch list, we never merge with it.
Fix this from the direct issue path by flagging the request as
REQ_NOMERGE so we don't change the size of it before issue.
See also:
https://bugzilla.kernel.org/show_bug.cgi?id=201685
Tested-by: Guenter Roeck <linux@roeck-us.net>
Fixes: 6ce3dd6eec1 ("blk-mq: issue directly if hw queue isn't busy in case of 'none'")
Cc: stable@vger.kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Adam W. Willis <return.of.octobot@gmail.com>
(cherry picked from commit 9a897ce1d5b6611daa27bf00fcfb5c97a3d826b4)
(cherry picked from commit 66af19f52cf6d2a9deef8de2f451604d49ef42f1)
In binder, using GFP_HIGHMEM will result in the allocated memory
not to be mapped in the kernel's virtual address space.
This prevents the kernel from being capable of directly
referring it.
Change-Id: I952dbc8ae205e47fa00ddf186ef306903f623367
Signed-off-by: Panchajanya1999 <panchajanya@azure-dev.live>
Signed-off-by: Jebaitedneko <Jebaitedneko@gmail.com>
According to Google we should set this to 0
as there is excessive logging in specific usecases
which has a negative impact on latency.
Signed-off-by: UtsavBalar1231 <utsavbalar1231@gmail.com>
Change-Id: Id619335848802e9d9a9bc13100d09a2cadbab07a
Hybrid mode could be switching between polling and interrupt mode. In
which case, we should always clear the pending IRQs to avoid spurious
interrupts.
Bug: 288490582
Test: Device boots, GCA, CTS
Signed-off-by: Edmond Chung <edmondchung@google.com>
Change-Id: Id33160b4c724cf800430c0833ce6703a5c2946ef
If the last CPU frequency selected isn't set before a new CPU frequency
selection arrives, then use the new selection immediately to avoid using a
stale frequency choice. This improves both performance and energy by more
closely tracking the scheduler's latest decisions.
Signed-off-by: Sultan Alsawaf <sultan@kerneltoast.com>
[Flopster101: Adapted to Exynos energy_aware governor]
Signed-off-by: Nahuel Gómez <nahuelgomez329@gmail.com>
The schedutil governor reduces frequencies too fast in some
situations which cases undesirable performance drops to
appear.
To address that issue, make schedutil reduce the frequency slower by
setting it to the average of the value chosen during the previous
iteration of governor computations and the new one coming from its
frequency selection formula.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=194963
Reported-by: John <john.ettedgui@gmail.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Cykeek <Cykeek@proton.me>
Signed-off-by: negrroo <mohammedaelnaggar1@gmail.com>
Signed-off-by: priiii1808 <priyanshusinghal0818@gmail.com>
[Flopster101: Adapted to Exynos energy_aware governor]
Signed-off-by: Nahuel Gómez <nahuelgomez329@gmail.com>
This is empirically observed to yield good performance with reduced power
consumption via having the down rate limit configured to be 2x longer than
the up rate limit. This reduces bouncing between CPU frequencies by
stalling down-clocking, which not only improves performance, but also
counter-intuitively improves power consumption.
The short up/down rate limits also provide improved interactivity and
real-time response.
Signed-off-by: Sultan Alsawaf <sultan@kerneltoast.com>
[Flopster101: Adapted to Exynos energy_aware governor]
Signed-off-by: Nahuel Gómez <nahuelgomez329@gmail.com>
Use tasklet_hi_schedule for better audio performance,
especially for LLA (Low Latency Audio) situation.
Signed-off-by: Sugar Zhang <sugar.zhang@rock-chips.com>
Change-Id: Ic5a215a269e718b0e5613132cb9fe9b58940d0e1
Testing:
[ElectroPerf & resist15]
In testing we found out that there were significant improvements
in the sequential read and write speeds. Some screenshots of the tests are below:
Before: https://i.imgur.com/UBL74X2.jpg
After: https://i.imgur.com/CrkD5iE.jpg
Change-Id: Idd7f5c7df0a7fc1535555927923491ecb39bc6a9
[Tashar02: Apply patch on kernel]
Signed-off-by: Tashfin Shakeer Rhythm <tashfinshakeerrhythm@gmail.com>
There's no reason to enter idle at this point in __CPU_PM_CPU_IDLE_ENTER()
if the CPU needs to reschedule. Instead of fruitlessly entering the
architecture's idle routine, reject the idle entry attempt with an error as
an optimization.
Signed-off-by: Sultan Alsawaf <sultan@kerneltoast.com>
Some architectures, such as arm, have implemented optimized copy_page for
full page copying.
Replace the full page memcpy with copy_page to take advantage of the
optimization.
Link: https://lkml.kernel.org/r/20231007070554.8657-1-mark-pk.tsai@mediatek.com
Signed-off-by: Mark-PK Tsai <mark-pk.tsai@mediatek.com>
Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Matthias Brugger <matthias.bgg@gmail.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: YJ Chiang <yj.chiang@mediatek.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Some architectures have implemented optimized copy_page for full page
copying, such as arm.
On my arm platform, use the copy_page helper for single page copying is
about 10 percent faster than memcpy.
Link: https://lkml.kernel.org/r/20231006060245.7411-1-mark-pk.tsai@mediatek.com
Signed-off-by: Mark-PK Tsai <mark-pk.tsai@mediatek.com>
Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Cc: Matthias Brugger <matthias.bgg@gmail.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: YJ Chiang <yj.chiang@mediatek.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
commit 3c12466b6b7bf1e56f9b32c366a3d83d87afb4de upstream.
Currently EROFS can map another compressed buffer for inplace
decompression, that was used to handle the cases that some pages of
compressed data are actually not in-place I/O.
However, like most simple LZ77 algorithms, LZ4 expects the compressed
data is arranged at the end of the decompressed buffer and it
explicitly uses memmove() to handle overlapping:
__________________________________________________________
|_ direction of decompression --> ____ |_ compressed data _|
Although EROFS arranges compressed data like this, it typically maps two
individual virtual buffers so the relative order is uncertain.
Previously, it was hardly observed since LZ4 only uses memmove() for
short overlapped literals and x86/arm64 memmove implementations seem to
completely cover it up and they don't have this issue. Juhyung reported
that EROFS data corruption can be found on a new Intel x86 processor.
After some analysis, it seems that recent x86 processors with the new
FSRM feature expose this issue with "rep movsb".
Let's strictly use the decompressed buffer for lz4 inplace
decompression for now. Later, as an useful improvement, we could try
to tie up these two buffers together in the correct order.
Reported-and-tested-by: Juhyung Park <qkrwngud825@gmail.com>
Closes: https://lore.kernel.org/r/CAD14+f2AVKf8Fa2OO1aAUdDNTDsVzzR6ctU_oJSmTyd6zSYR2Q@mail.gmail.com
Fixes: 0ffd71bcc3a0 ("staging: erofs: introduce LZ4 decompression inplace")
Fixes: 598162d05080 ("erofs: support decompress big pcluster for lz4 backend")
Cc: stable <stable@vger.kernel.org> # 5.4+
Tested-by: Yifan Zhao <zhaoyifan@sjtu.edu.cn>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20231206045534.3920847-1-hsiangkao@linux.alibaba.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The msm_irqbalance service in userspace constantly migrates IRQs between
CPUs according to its whims, which is not desired. All of the IRQs have
a sane affinity (CPU0 if unimportant, CPU4-7 otherwise), so prevent
userspace from tampering with that.
Signed-off-by: Sultan Alsawaf <sultanxda@gmail.com>
(cherry picked from commit 6cedf3c9b1f8c962d19ce4151ca5caaff69e3c6a)
(cherry picked from commit 8fc0013ba4094fd8fe95fb0d23af0936347060f4)
We this to access task_is_booster().
ld.lld: error: undefined symbol: task_is_booster
>>> referenced by elevator.c:774 (../block/elevator.c:774)
>>> vmlinux.o:(elv_iosched_store)
>>> did you mean: task_is_booster
>>> defined in: vmlinux.o
Signed-off-by: Nahuel Gómez <nahuelgomez329@gmail.com>
I'm not sure if this will help, but the idea is to give the codec more room for error, since currently there is audio crackling under moderate CPU load.
Signed-off-by: Nahuel Gómez <nahuelgomez329@gmail.com>
This patch fixes the latency_sensitive flag for all cpuset cgroups, and
the value present in the uclamp.latency_sensitive node directly
corresponds to the task_group's latency_sensitive value.
Prior to this patch, this was not the case. The
uclamp_latency_sensitive() function applied values only to the cpu
cgroup subsys instead of the required cpuset cgroup subsys, as a
result of which the latency_sensitive value remained zero for all
taskgroups irrespective of its setting.
Also, fix a situation where latency_sensitive is enabled for the
cpuset's root cgroup, in which case all tasks will have their value
as 1, which in turn will enable prefer_idle for all tasks. This is
undesired and may cause high battery drain.
Earlier commits in this series allow battery-powered systems to build
their kernels with the default-disabled CONFIG_RCU_LAZY=y Kconfig option.
This Kconfig option causes call_rcu() to delay its callbacks in order
to batch them. This means that a given RCU grace period covers more
callbacks, thus reducing the number of grace periods, in turn reducing
the amount of energy consumed, which increases battery lifetime which
can be a very good thing. This is not a subtle effect: In some important
use cases, the battery lifetime is increased by more than 10%.
This CONFIG_RCU_LAZY=y option is available only for CPUs that offload
callbacks, for example, CPUs mentioned in the rcu_nocbs kernel boot
parameter passed to kernels built with CONFIG_RCU_NOCB_CPU=y.
Delaying callbacks is normally not a problem because most callbacks do
nothing but free memory. If the system is short on memory, a shrinker
will kick all currently queued lazy callbacks out of their laziness,
thus freeing their memory in short order. Similarly, the rcu_barrier()
function, which blocks until all currently queued callbacks are invoked,
will also kick lazy callbacks, thus enabling rcu_barrier() to complete
in a timely manner.
However, there are some cases where laziness is not a good option.
For example, synchronize_rcu() invokes call_rcu(), and blocks until
the newly queued callback is invoked. It would not be a good for
synchronize_rcu() to block for ten seconds, even on an idle system.
Therefore, synchronize_rcu() invokes call_rcu_flush() instead of
call_rcu(). The arrival of a non-lazy call_rcu_flush() callback on a
given CPU kicks any lazy callbacks that might be already queued on that
CPU. After all, if there is going to be a grace period, all callbacks
might as well get full benefit from it.
Yes, this could be done the other way around by creating a
call_rcu_lazy(), but earlier experience with this approach and
feedback at the 2022 Linux Plumbers Conference shifted the approach
to call_rcu() being lazy with call_rcu_flush() for the few places
where laziness is inappropriate.
And another call_rcu() instance that cannot be lazy is the one
in queue_rcu_work(), given that callers to queue_rcu_work() are
not necessarily OK with long delays.
Therefore, make queue_rcu_work() use call_rcu_flush() in order to revert
to the old behavior.
Signed-off-by: Uladzislau Rezki <urezki@gmail.com>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Simultaneously improves interactivity and power efficiency
[Flopster101: This also invalidates any value set by RCU_BOOST_DELAY.]
Signed-off-by: Nahuel Gómez <nahuelgomez329@gmail.com>
* Fixes:
drivers/gpu/arm/bv_r38p1/Kconfig:389:warning: ignoring type redefinition of 'MALI_ARBITRATION' from 'bool' to 'tristate'
Change-Id: Ia7a7d1a4fd68344abd1c07f7bb5e9ef214bdc51c
[ Upstream commit cc9b364bb1d58d3dae270c7a931a8cc717dc2b3b ]
There are race conditions that may lead to inet6_dev refcount underflow
in xfrm6_dst_destroy() and rt6_uncached_list_flush_dev().
One of the refcount underflow bugs is shown below:
(cpu 1) | (cpu 2)
xfrm6_dst_destroy() |
... |
in6_dev_put() |
| rt6_uncached_list_flush_dev()
... | ...
| in6_dev_put()
rt6_uncached_list_del() | ...
... |
xfrm6_dst_destroy() calls rt6_uncached_list_del() after in6_dev_put(),
so rt6_uncached_list_flush_dev() has a chance to call in6_dev_put()
again for the same inet6_dev.
Fix it by moving in6_dev_put() after rt6_uncached_list_del() in
xfrm6_dst_destroy().
Fixes: 510c321b5571 ("xfrm: reuse uncached_list to track xdsts")
Signed-off-by: Zhang Changzhong <zhangchangzhong@huawei.com>
Reviewed-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit cb3871b1cd135a6662b732fbc6b3db4afcdb4a64 upstream.
The code pattern of memcpy(dst, src, strlen(src)) is almost always
wrong. In this case it is wrong because it leaves memory uninitialized
if it is less than sizeof(ni->name), and overflows ni->name when longer.
Normally strtomem_pad() could be used here, but since ni->name is a
trailing array in struct hci_mon_new_index, compilers that don't support
-fstrict-flex-arrays=3 can't tell how large this array is via
__builtin_object_size(). Instead, open-code the helper and use sizeof()
since it will work correctly.
Additionally mark ni->name as __nonstring since it appears to not be a
%NUL terminated C string.
Cc: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Cc: Edward AD <twuufnxlz@gmail.com>
Cc: Marcel Holtmann <marcel@holtmann.org>
Cc: Johan Hedberg <johan.hedberg@gmail.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: linux-bluetooth@vger.kernel.org
Cc: netdev@vger.kernel.org
Fixes: 18f547f3fc07 ("Bluetooth: hci_sock: fix slab oob read in create_monitor_event")
Link: https://lore.kernel.org/lkml/202310110908.F2639D3276@keescook/
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 18f547f3fc074500ab5d419cf482240324e73a7e upstream.
When accessing hdev->name, the actual string length should prevail
Reported-by: syzbot+c90849c50ed209d77689@syzkaller.appspotmail.com
Fixes: dcda165706b9 ("Bluetooth: hci_core: Fix build warnings")
Signed-off-by: Edward AD <twuufnxlz@gmail.com>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 3b384cc74b00b5ac21d18e4c1efc3c1da5300971 ]
Looks like the driver sleep pins configuration is unusable. Adding the
sleep pins causes the usb phy to not respond. We need to use the default
pins in probe, and only set sleep pins at phy_mdm6600_device_power_off().
As the modem can also be booted to a serial port mode for firmware
flashing, let's make the pin changes limited to probe and remove. For
probe, we get the default pins automatically. We only need to set the
sleep pins in phy_mdm6600_device_power_off() to prevent the modem from
waking up because the gpio line glitches.
If it turns out that we need a separate state for phy_mdm6600_power_on()
and phy_mdm6600_power_off(), we can use the pinctrl idle state.
Cc: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>
Cc: Merlijn Wajer <merlijn@wizzup.org>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Sebastian Reichel <sre@kernel.org>
Fixes: 2ad2af081622 ("phy: mapphone-mdm6600: Improve phy related runtime PM calls")
Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com>
Link: https://lore.kernel.org/r/20230913060433.48373-3-tony@atomide.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b99e0ba9633af51638e5ee1668da2e33620c134f ]
Otherwise we will get an underflow on remove.
Cc: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>
Cc: Merlijn Wajer <merlijn@wizzup.org>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Sebastian Reichel <sre@kernel.org>
Fixes: f7f50b2a7b05 ("phy: mapphone-mdm6600: Add runtime PM support for n_gsm on USB suspend")
Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com>
Link: https://lore.kernel.org/r/20230913060433.48373-2-tony@atomide.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 719606154c7033c068a5d4c1dc5f9163b814b3c8 ]
Commit d644e0d79829 ("phy: mapphone-mdm6600: Fix PM error handling in
phy_mdm6600_probe") caused a regression where we now unconditionally
disable runtime PM at the end of the probe while it is only needed on
errors.
Cc: Ivaylo Dimitrov <ivo.g.dimitrov.75@gmail.com>
Cc: Merlijn Wajer <merlijn@wizzup.org>
Cc: Miaoqian Lin <linmq006@gmail.com>
Cc: Pavel Machek <pavel@ucw.cz>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com>
Fixes: d644e0d79829 ("phy: mapphone-mdm6600: Fix PM error handling in phy_mdm6600_probe")
Signed-off-by: Tony Lindgren <tony@atomide.com>
Link: https://lore.kernel.org/r/20230913060433.48373-1-tony@atomide.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit aa6464edbd51af4a2f8db43df866a7642b244b5f ]
Free the "priv" pointer before returning the error code.
Fixes: 90eb6b59d311 ("ASoC: pxa-ssp: add support for an external clock in devicetree")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://lore.kernel.org/r/84ac2313-1420-471a-b2cb-3269a2e12a7c@moroto.mountain
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit fc363413ef8ea842ae7a99e3caf5465dafdd3a49 upstream.
We found a glitch when configuring the pad as output high. To avoid this
glitch, move the data value setting before direction config in the
function vf610_gpio_direction_output().
Fixes: 659d8a62311f ("gpio: vf610: add imx7ulp support")
Signed-off-by: Haibo Chen <haibo.chen@nxp.com>
[Bartosz: tweak the commit message]
Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 235985d1763f7aba92c1c64e5f5aaec26c2c9b18 upstream.
Newer Asus laptops send the following new WMI event codes when some
of the F1 - F12 "media" hotkeys are pressed:
0x2a Screen Capture
0x2b PrintScreen
0x2c CapsLock
Map 0x2a to KEY_SELECTIVE_SCREENSHOT mirroring how similar hotkeys
are mapped on other laptops.
PrintScreem and CapsLock are also reported as normal PS/2 keyboard events,
map these event codes to KE_IGNORE to avoid "Unknown key code 0x%x\n" log
messages.
Reported-by: James John <me@donjajo.com>
Closes: https://lore.kernel.org/platform-driver-x86/a2c441fe-457e-44cf-a146-0ecd86b037cf@donjajo.com/
Closes: https://bbs.archlinux.org/viewtopic.php?pid=2123716
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20231017090725.38163-4-hdegoede@redhat.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f37cc2fc277b371fc491890afb7d8a26e36bb3a1 upstream.
Older Asus laptops change the backlight level themselves and then send
WMI events with different codes for different backlight levels.
The asus-wmi.c code maps the entire range of codes reported on
brightness down keypresses to an internal ASUS_WMI_BRN_DOWN code:
define NOTIFY_BRNUP_MIN 0x11
define NOTIFY_BRNUP_MAX 0x1f
define NOTIFY_BRNDOWN_MIN 0x20
define NOTIFY_BRNDOWN_MAX 0x2e
if (code >= NOTIFY_BRNUP_MIN && code <= NOTIFY_BRNUP_MAX)
code = ASUS_WMI_BRN_UP;
else if (code >= NOTIFY_BRNDOWN_MIN && code <= NOTIFY_BRNDOWN_MAX)
code = ASUS_WMI_BRN_DOWN;
Before this commit all the NOTIFY_BRNDOWN_MIN - NOTIFY_BRNDOWN_MAX
aka 0x20 - 0x2e events were mapped to 0x20.
This mapping is causing issues on new laptop models which actually
send 0x2b events for printscreen presses and 0x2c events for
capslock presses, which get translated into spurious brightness-down
presses.
The plan is disable the 0x11-0x2e special mapping on laptops
where asus-wmi does not register a backlight-device to avoid
the spurious brightness-down keypresses. New laptops always send
0x2e for brightness-down presses, change the special internal
ASUS_WMI_BRN_DOWN value from 0x20 to 0x2e to match this in
preparation for fixing the spurious brightness-down presses.
This change does not have any functional impact since all
of 0x20 - 0x2e is mapped to ASUS_WMI_BRN_DOWN first and only
then checked against the keymap code and the new 0x2e
value is still in the 0x20 - 0x2e range.
Reported-by: James John <me@donjajo.com>
Closes: https://lore.kernel.org/platform-driver-x86/a2c441fe-457e-44cf-a146-0ecd86b037cf@donjajo.com/
Closes: https://bbs.archlinux.org/viewtopic.php?pid=2123716
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20231017090725.38163-2-hdegoede@redhat.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c1ae1c59c8c6e0b66a718308c623e0cb394dab6b upstream.
Since the fixed commits both zdev->iommu_bitmap and zdev->lazy_bitmap
are allocated as vzalloc(zdev->iommu_pages / 8). The problem is that
zdev->iommu_bitmap is a pointer to unsigned long but the above only
yields an allocation that is a multiple of sizeof(unsigned long) which
is 8 on s390x if the number of IOMMU pages is a multiple of 64.
This in turn is the case only if the effective IOMMU aperture is
a multiple of 64 * 4K = 256K. This is usually the case and so didn't
cause visible issues since both the virt_to_phys(high_memory) reduced
limit and hardware limits use nice numbers.
Under KVM, and in particular with QEMU limiting the IOMMU aperture to
the vfio DMA limit (default 65535), it is possible for the reported
aperture not to be a multiple of 256K however. In this case we end up
with an iommu_bitmap whose allocation is not a multiple of
8 causing bitmap operations to access it out of bounds.
Sadly we can't just fix this in the obvious way and use bitmap_zalloc()
because for large RAM systems (tested on 8 TiB) the zdev->iommu_bitmap
grows too large for kmalloc(). So add our own bitmap_vzalloc() wrapper.
This might be a candidate for common code, but this area of code will
be replaced by the upcoming conversion to use the common code DMA API on
s390 so just add a local routine.
Fixes: 224593215525 ("s390/pci: use virtual memory for iommu bitmap")
Fixes: 13954fd6913a ("s390/pci_dma: improve lazy flush for unmap")
Cc: stable@vger.kernel.org
Reviewed-by: Matthew Rosato <mjrosato@linux.ibm.com>
Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
