If FEAT_MTE2 is disabled via the arm64.nomte command line argument on a
CPU that claims to support FEAT_MTE2, the kernel will use Tagged Normal
in the MAIR. If we interpret arm64.nomte to mean that the CPU does not
in fact implement FEAT_MTE2, setting the system register like this may
lead to UNSPECIFIED behavior. Fix it by arranging for MAIR to be set
in the C function cpu_enable_mte which is called based on the sanitized
version of the system register.
There is no need for the rest of the MTE-related system register
initialization to happen from assembly, with the exception of TCR_EL1,
which must be set to include at least TBI1 because the secondary CPUs
access KASan-allocated data structures early. Therefore, make the TCR_EL1
initialization unconditional and move the rest of the initialization to
cpu_enable_mte so that we no longer have a dependency on the unsanitized
ID register value.
Signed-off-by: Peter Collingbourne <pcc@google.com>
Signed-off-by: Evgenii Stepanov <eugenis@google.com>
Signed-off-by: Qun-Wei Lin <qun-wei.lin@mediatek.corp-partner.google.com>
Suggested-by: Catalin Marinas <catalin.marinas@arm.com>
Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/linux-arm-kernel/20220907003630.1115439-1-eugenis@google.com/
Bug: 239834217
Change-Id: I7bd247ae5802295c198bac42df86fab9e9e9cefd
(cherry picked from commit 6a2905d35289caeeecbafa93ae4ff7cefafd02ed)
Remove the __read_mostly attributes from the rcu_scheduler_active
extern declarations, because these attributes are ignored for
prototypes and we'd have to include the full <linux/cache.h> header
to gain this functionally pointless attribute defined.
BUG: 274028827
Change-Id: Ib9c1d34ad2325b8cf939f5df0b35ebe4b281e9ec
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
(cherry picked from commit e6339d3b44)
Signed-off-by: Srinivasarao Pathipati <quic_spathi@quicinc.com>
Currently we process the suspend interrupt event only if the
device is in configured state. Consider a case where device
is not configured and got suspend interrupt, in that case our
gadget will still use 100mA as composite_suspend didn't happen.
But battery charging specification (BC1.2) expects a downstream
device to draw less than 2.5mA when unconnected OR suspended.
Fix this by removing the condition for processing suspend event,
and thus composite_resume would set vbus draw to 2.
Fixes: 72704f876f ("dwc3: gadget: Implement the suspend entry event handler")
Signed-off-by: Prashanth K <quic_prashk@quicinc.com>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://lore.kernel.org/r/1677217619-10261-2-git-send-email-quic_prashk@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Bug: 272680990
(cherry picked from commit 4decf4060ehttps://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb.git/ usb-next)
Change-Id: I902b8911debb7c049a4a0128077ea8d931900d1d
Signed-off-by: Prashanth K <quic_prashk@quicinc.com>
commit ee059170b1 upstream.
The imperfect hash area can be updated while packets are traversing,
which will cause a use-after-free when 'tcf_exts_exec()' is called
with the destroyed tcf_ext.
CPU 0: CPU 1:
tcindex_set_parms tcindex_classify
tcindex_lookup
tcindex_lookup
tcf_exts_change
tcf_exts_exec [UAF]
Stop operating on the shared area directly, by using a local copy,
and update the filter with 'rcu_replace_pointer()'. Delete the old
filter version only after a rcu grace period elapsed.
Bug: 272274675
Fixes: 9b0d4446b5 ("net: sched: avoid atomic swap in tcf_exts_change")
Reported-by: valis <sec@valis.email>
Suggested-by: valis <sec@valis.email>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Pedro Tammela <pctammela@mojatatu.com>
Link: https://lore.kernel.org/r/20230209143739.279867-1-pctammela@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Lee Jones <joneslee@google.com>
Change-Id: I02ce525a338242efdb8080e2e2c705d3b59d9a68
Ignore kvm-arm.mode if !is_hyp_mode_available(). Specifically, we want
to avoid switching kvm_mode to KVM_MODE_PROTECTED if hypervisor mode is
not available. This prevents "Protected KVM" cpu capability being
reported when Linux is booting in EL1 and would not have KVM enabled.
Reasonably though, we should warn if the command line is requesting a
KVM mode at all if KVM isn't actually available. Allow
"kvm-arm.mode=none" to skip the warning since this would disable KVM
anyway.
Signed-off-by: Elliot Berman <quic_eberman@quicinc.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20220920190658.2880184-1-quic_eberman@quicinc.com
Bug: 249052880
(cherry picked from commit b2a4d007c3https://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm.git next)
Change-Id: I196649fc00537e563a1fab0a22bf23c5b7abe00d
Signed-off-by: Elliot Berman <quic_eberman@quicinc.com>
(cherry picked from commit 8015dd49c0)
Although KVM can be compiled out of the kernel, it cannot be disabled
at runtime. Allow this possibility by introducing a new mode that
will prevent KVM from initialising.
This is useful in the (limited) circumstances where you don't want
KVM to be available (what is wrong with you?), or when you want
to install another hypervisor instead (good luck with that).
Reviewed-by: David Brazdil <dbrazdil@google.com>
Acked-by: Will Deacon <will@kernel.org>
Acked-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Reviewed-by: Andrew Scull <ascull@google.com>
Link: https://lore.kernel.org/r/20211001170553.3062988-1-maz@kernel.org
(cherry picked from commit b6a68b97af)
Bug: 204960018
Signed-off-by: Will Deacon <willdeacon@google.com>
Change-Id: Ie796c716fc7cece906a8cded0ae4652a828988bb
(cherry picked from commit a621a4e6cb)
The cuttlefish symbol list is out of date. Updating it with
the following steps.
$ BUILD_CONFIG=common/build.config.gki.aarch64 build/build.sh
$ BUILD_CONFIG=common-modules/virtual-\
device/build.config.cuttlefish.aarch64 build/build.sh
$ build/abi/extract_symbols out/android12-5.10/dist/ --whitelist \
common/android/abi_gki_aarch64_virtual_device --additions-only
$ BUILD_CONFIG=common/build.config.gki.aarch64 build/build_abi.sh \
--update --print-report
Bug: 203806616
Test: Local boot on cuttlefish arm64 android11-gsi
Change-Id: I48fbc598e5a3e4ac8db68c51d699922099da7de8
Signed-off-by: Ram Muthiah <rammuthiah@google.com>
This reverts commit 9d2ec2e0b6.
The hook android_rvh_setup_dma_ops is not used by any vendor, so remove
it to help with merge issues with future LTS releases.
If this is needed by any real user, it can easily be reverted to add it
back and then the symbol should be added to the abi list at the same
time to prevent it from being removed again later.
Bug: 263236925
Bug: 203756332
Cc: Beata Michalska <beata.michalska@arm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I7741910039ac9914e2e2adf974204278d37db94d
We can't use 0 here, as io_init_req() is always invoked with the
ctx uring_lock held. Newer kernels have IO_URING_F_UNLOCKED for this,
but previously we used IO_URING_F_NONBLOCK to indicate this as well.
Fixes: 08681391b8 ("io_uring: add missing lock in io_get_file_fixed")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
(cherry picked from commit da24142b1e)
Bug: 268174392
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I762eacf1b49ca8a38d8b77c44db4ca2bc49b2c4c
During enumeration or composition switch,a userspace process
agnostic of the conventions of configs can try to create function
symlinks even after the UDC is bound to current config which is
not correct. Potentially it can create duplicates within the
current config.
Prevent this by adding a check if udc_name already exists, then bail
out of cfg_link.
Following is an example:
Step1:
ln -s X1 ffs.a
-->cfg_link
--> usb_get_function(ffs.a)
->ffs_alloc
CFG->FUNC_LIST: <ffs.a>
C->FUNCTION: <empty>
Step2:
echo udc.name > /config/usb_gadget/g1/UDC
--> UDC_store
->composite_bind
->usb_add_function
CFG->FUNC_LIST: <empty>
C->FUNCTION: <ffs.a>
Step3:
ln -s Y1 ffs.a
-->cfg_link
-->usb_get_function(ffs.a)
->ffs_alloc
CFG->FUNC_LIST: <ffs.a>
C->FUNCTION: <ffs.a>
both the lists corresponds to the same function instance ffs.a
but the usb_function* pointer is different because in step 3
ffs_alloc has created a new reference to usb_function* for
ffs.a and added it to cfg_list.
Step4:
Now a composition switch involving <ffs.b,ffs.a> is executed.
the composition switch will involve 3 things:
1. unlinking the previous functions existing
2. creating new symlinks
3. writing UDC
However, the composition switch is generally taken care by
userspace process which creates the symlinks in its own
nomenclature(X*) and removes only those.
So it won't be able to remove Y1 which user had created
by own.
Due to this the new symlinks cannot be created for ffs.a
since the entry already exists in CFG->FUNC_LIST.
The state of the CFG->FUNC_LIST is as follows:
CFG->FUNC_LIST: <ffs.a>
Fixes: 88af8bbe4e ("usb: gadget: the start of the configfs interface")
Signed-off-by: Krishna Kurapati PSSNV <quic_kriskura@quicinc.com>
Signed-off-by: Udipto Goswami <quic_ugoswami@quicinc.com>
Link: https://lore.kernel.org/r/20230201132308.31523-1-quic_ugoswami@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
BUG: 267715065
(cherry picked from commit 89e7252d6c usb-next)
Change-Id: I0f21c75ce0929756efacd0035ae21085ab9742f2
Signed-off-by: Udipto Goswami <quic_ugoswami@quicinc.com>
io_get_file_fixed will access io_uring's context. Lock it if it is
invoked unlocked (eg via io-wq) to avoid a race condition with fixed
files getting unregistered.
No single upstream patch exists for this issue, it was fixed as part
of the file assignment changes that went into the 5.18 cycle.
Signed-off-by: Jheng, Bing-Jhong Billy <billy@starlabs.sg>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
(cherry picked from commit 08681391b8)
Bug: 268174392
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Change-Id: I01ec4283589acde1d17318eb76f87ce099ec3fa0
These hooks help us do the following things:
a) Record the number of mutex and rwsem optimistic spin.
b) Monitor the time of mutex and rwsem optimistic spin.
c) Make it possible if oems don't want mutex and rwsem to optimistic spin
for a long time.
Bug: 267565260
Change-Id: I2bee30fb17946be85e026213b481aeaeaee2459f
Signed-off-by: Liujie Xie <xieliujie@oppo.com>
On backports where the dmabuf_page_pool mutex could not be completely
removed because of KMI constraints, we lock the mutex to prevent anyone
from acquiring it. While this is fine for regular builds, lockdep
reports a deadlock because of this behavior. So here we remove the call
to mutex_lock for this unused mutex.
Bug: 269576779
Fixes: d55aeb4029 ("BACKPORT: ANDROID: dma-buf: heaps: replace mutex lock with spinlock")
Change-Id: Ic426b6e435ec4ac7b0355c50ec9ea14d38acfa43
Signed-off-by: T.J. Mercier <tjmercier@google.com>
find_get_page() returns a page with increased refcount, assuming a page
exists at the given index. Ensure this refcount is dropped on error.
Bug: 271079833
Fixes: 59d4d125 ("BACKPORT: FROMLIST: mm: implement speculative handling in filemap_fault()")
Change-Id: Idc7b9e3f11f32a02bed4c6f4e11cec9200a5c790
Signed-off-by: Patrick Daly <quic_pdaly@quicinc.com>
(cherry picked from commit 6232eecfa7ca0d8d0ca088da6d0edb2c3a879ff9)
Signed-off-by: Zhenhua Huang <quic_zhenhuah@quicinc.com>
[ Upstream commit 2e7eab8142 ]
According to Intel's document on Indirect Branch Restricted
Speculation, "Enabling IBRS does not prevent software from controlling
the predicted targets of indirect branches of unrelated software
executed later at the same predictor mode (for example, between two
different user applications, or two different virtual machines). Such
isolation can be ensured through use of the Indirect Branch Predictor
Barrier (IBPB) command." This applies to both basic and enhanced IBRS.
Since L1 and L2 VMs share hardware predictor modes (guest-user and
guest-kernel), hardware IBRS is not sufficient to virtualize
IBRS. (The way that basic IBRS is implemented on pre-eIBRS parts,
hardware IBRS is actually sufficient in practice, even though it isn't
sufficient architecturally.)
For virtual CPUs that support IBRS, add an indirect branch prediction
barrier on emulated VM-exit, to ensure that the predicted targets of
indirect branches executed in L1 cannot be controlled by software that
was executed in L2.
Since we typically don't intercept guest writes to IA32_SPEC_CTRL,
perform the IBPB at emulated VM-exit regardless of the current
IA32_SPEC_CTRL.IBRS value, even though the IBPB could technically be
deferred until L1 sets IA32_SPEC_CTRL.IBRS, if IA32_SPEC_CTRL.IBRS is
clear at emulated VM-exit.
This is CVE-2022-2196.
Bug: 265479180
Fixes: 5c911beff2 ("KVM: nVMX: Skip IBPB when switching between vmcs01 and vmcs02")
Cc: Sean Christopherson <seanjc@google.com>
Signed-off-by: Jim Mattson <jmattson@google.com>
Reviewed-by: Sean Christopherson <seanjc@google.com>
Link: https://lore.kernel.org/r/20221019213620.1953281-3-jmattson@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Lee Jones <joneslee@google.com>
Change-Id: Ic3ac006d834719d41f8a84c7a557d60b82e05197
Fixing below build error that got exposed by 'commit 824c55581d
("UPSTREAM: Enable '-Werror' by default for all kernel builds")'
by making show_rcu_tasks_rude_gp_kthread() function as 'inline'.
In file included from kernel/rcu/update.c:579:0:
kernel/rcu/tasks.h:710:13: error: ‘show_rcu_tasks_rude_gp_kthread’ defined but not used [-Werror=unused-function]
static void show_rcu_tasks_rude_gp_kthread(void) {}
All warnings are treated as errors after picking upstream
'commit 824c55581d ("UPSTREAM: Enable '-Werror' by default
for all kernel builds")' into android[12/13]-5.10 kernels.
In latest Linux kernels this issue won't exist as problematic code got
modified by 'commit 27c0f14483 ("rcutorture: Make grace-period kthread
report match RCU flavor being tested")'.
Even though buggy code present in 5.10 stable kernel, the
'commit 824c55581d ("UPSTREAM: Enable '-Werror' by default for all
kernel builds")' is not present so this fix is not relevant here
and got rejected by community.
Bug: 269057599
Link: https://lore.kernel.org/stable/1676916839-32235-1-git-send-email-quic_c_spathi@quicinc.com/
Fixes: 824c55581d ("UPSTREAM: Enable '-Werror' by default for all kernel builds")
Change-Id: I95f6a7a81883d7a6b0f56068b1eb49dc02788685
Signed-off-by: Srinivasarao Pathipati <quic_spathi@quicinc.com>
Bug: 241479010
Test: incfs_test passes, play confirm behavior in bug is fixed
Signed-off-by: Paul Lawrence <paullawrence@google.com>
Change-Id: Ie51f2b76d0873057f54fecf7fcc793c66df20969
Also fix race whereby multiple providers writinig the same block would
actually write out the same block.
Note that multiple_providers_test started failing when incfs was ported
to 5.15, and these fixes are needed to make the test reliable
Bug: 264703896
Test: incfs-test passes, specifically multiple_providers_test. Ran 100
times
Change-Id: I05ad5b2b2f62cf218256222cecb79bbe9953bd97
Signed-off-by: Paul Lawrence <paullawrence@google.com>
This reverts commit 64ea6f3273.
The symbol get_avenrun is not used by any external modules, so remove
the export as it is useless. If this is needed, the original commit can
be added back and the symbol added to the relevant symbol list.
Bug: 158067689
Bug: 203756332
Cc: Abhilasha Rao <abhilasha.hv@samsung.corp-partner.google.com>
Cc: Todd Kjos <tkjos@google.com>
Change-Id: I394d074a6cd4afaaf8eb97d528e039bc168e2fe1
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit eee22187b5 upstream.
In do_writepages, if the value returned by ext4_writepages is "-ENOMEM"
and "wbc->sync_mode == WB_SYNC_ALL", retry until the condition is not met.
In __ext4_get_inode_loc, if the bh returned by sb_getblk is NULL,
the function returns -ENOMEM.
In __getblk_slow, if the return value of grow_buffers is less than 0,
the function returns NULL.
When the three processes are connected in series like the following stack,
an infinite loop may occur:
do_writepages <--- keep retrying
ext4_writepages
mpage_map_and_submit_extent
mpage_map_one_extent
ext4_map_blocks
ext4_ext_map_blocks
ext4_ext_handle_unwritten_extents
ext4_ext_convert_to_initialized
ext4_split_extent
ext4_split_extent_at
__ext4_ext_dirty
__ext4_mark_inode_dirty
ext4_reserve_inode_write
ext4_get_inode_loc
__ext4_get_inode_loc <--- return -ENOMEM
sb_getblk
__getblk_gfp
__getblk_slow <--- return NULL
grow_buffers
grow_dev_page <--- return -ENXIO
ret = (block < end_block) ? 1 : -ENXIO;
In this issue, bg_inode_table_hi is overwritten as an incorrect value.
As a result, `block < end_block` cannot be met in grow_dev_page.
Therefore, __ext4_get_inode_loc always returns '-ENOMEM' and do_writepages
keeps retrying. As a result, the writeback process is in the D state due
to an infinite loop.
Add a check on inode table block in the __ext4_get_inode_loc function by
referring to ext4_read_inode_bitmap to avoid this infinite loop.
Cc: stable@kernel.org
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Ritesh Harjani (IBM) <ritesh.list@gmail.com>
Link: https://lore.kernel.org/r/20220817132701.3015912-3-libaokun1@huawei.com
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Bug: 264629802
Bug: 264632463
Change-Id: Id3bb71336059cac33f16fca383e783add3a01295
Signed-off-by: Tudor Ambarus <tudor.ambarus@linaro.org>
commit 5c61795ea9 upstream.
This debug statement was never meant to go into the upstream release,
kill it off before it ends up in a release. It was just part of the
testing for the initial version of the patch.
Fixes: 2ec33a6c3c ("io_uring/rw: ensure kiocb_end_write() is always called")
Change-Id: Iee9f436c34cc137a7ab934aafa3aa0c584369418
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Bug: 268174392
(cherry picked from commit e699cce29a)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit 2ec33a6c3c upstream.
A previous commit moved the notifications and end-write handling, but
it is now missing a few spots where we also want to call both of those.
Without that, we can potentially be missing file notifications, and
more importantly, have an imbalance in the super_block writers sem
accounting.
Fixes: b000145e99 ("io_uring/rw: defer fsnotify calls to task context")
Reported-by: Dave Chinner <david@fromorbit.com>
Link: https://lore.kernel.org/all/20221010050319.GC2703033@dread.disaster.area/
Change-Id: Iaaa509f5dadcae04f58c929901225bc968b35d52
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Bug: 268174392
(cherry picked from commit 3d5f181bda)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit 3e4cb6ebbb upstream.
I hit a very bad problem during my tests of SENDMSG_ZC.
BUG(); in first_iovec_segment() triggered very easily.
The problem was io_setup_async_msg() in the partial retry case,
which seems to happen more often with _ZC.
iov_iter_iovec_advance() may change i->iov in order to have i->iov_offset
being only relative to the first element.
Which means kmsg->msg.msg_iter.iov is no longer the
same as kmsg->fast_iov.
But this would rewind the copy to be the start of
async_msg->fast_iov, which means the internal
state of sync_msg->msg.msg_iter is inconsitent.
I tested with 5 vectors with length like this 4, 0, 64, 20, 8388608
and got a short writes with:
- ret=2675244 min_ret=8388692 => remaining 5713448 sr->done_io=2675244
- ret=-EAGAIN => io_uring_poll_arm
- ret=4911225 min_ret=5713448 => remaining 802223 sr->done_io=7586469
- ret=-EAGAIN => io_uring_poll_arm
- ret=802223 min_ret=802223 => res=8388692
While this was easily triggered with SENDMSG_ZC (queued for 6.1),
it was a potential problem starting with 7ba89d2af1
in 5.18 for IORING_OP_RECVMSG.
And also with 4c3c09439c in 5.19
for IORING_OP_SENDMSG.
However 257e84a537 introduced the critical
code into io_setup_async_msg() in 5.11.
Fixes: 7ba89d2af1 ("io_uring: ensure recv and recvmsg handle MSG_WAITALL correctly")
Fixes: 257e84a537 ("io_uring: refactor sendmsg/recvmsg iov managing")
Cc: stable@vger.kernel.org
Change-Id: I72c459fdbae2938d176126ed2f17eea990c42d49
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/b2e7be246e2fb173520862b0c7098e55767567a2.1664436949.git.metze@samba.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Bug: 268174392
(cherry picked from commit fc2491562a)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit 6f83ab22ad upstream.
-1 tells use to use the current position, but we check if the file is
a stream regardless of that. Fix up io_kiocb_update_pos() to only
dip into file if we need to. This is both more efficient and also drops
12 bytes of text on aarch64 and 64 bytes on x86-64.
Fixes: b4aec40015 ("io_uring: do not recalculate ppos unnecessarily")
Change-Id: I5c22ce8122b0e1f0ad423a5b3aa520ee416feff1
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Bug: 268174392
(cherry picked from commit 89a77271d2)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit b4aec40015 upstream.
There is a slight optimisation to be had by calculating the correct pos
pointer inside io_kiocb_update_pos and then using that later.
It seems code size drops by a bit:
000000000000a1b0 0000000000000400 t io_read
000000000000a5b0 0000000000000319 t io_write
vs
000000000000a1b0 00000000000003f6 t io_read
000000000000a5b0 0000000000000310 t io_write
Change-Id: I19d8cdb6ea88d8fc4625e521363d5a8f638dfdcb
Signed-off-by: Dylan Yudaken <dylany@fb.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bug: 268174392
(cherry picked from commit e90cfb9699)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit d34e1e5b39 upstream.
Update kiocb->ki_pos at execution time rather than in io_prep_rw().
io_prep_rw() happens before the job is enqueued to a worker and so the
offset might be read multiple times before being executed once.
Ensures that the file position in a set of _linked_ SQEs will be only
obtained after earlier SQEs have completed, and so will include their
incremented file position.
Change-Id: I3c5abbf6a337ec1958fd6600c5feb44fb61a5772
Signed-off-by: Dylan Yudaken <dylany@fb.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bug: 268174392
(cherry picked from commit ea528ecac3)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit af9c45eceb upstream.
io_kiocb_ppos is called in both branches, and it seems that the compiler
does not fuse this. Fusing removes a few bytes from loop_rw_iter.
Before:
$ nm -S fs/io_uring.o | grep loop_rw_iter
0000000000002430 0000000000000124 t loop_rw_iter
After:
$ nm -S fs/io_uring.o | grep loop_rw_iter
0000000000002430 000000000000010d t loop_rw_iter
Change-Id: Ibd662d59697d9cb1e484319050f6e5f960f6ac5c
Signed-off-by: Dylan Yudaken <dylany@fb.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bug: 268174392
(cherry picked from commit 076f872314)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
commit e775f93f2a upstream.
io_uring caches task references to avoid doing atomics for each of them
per request. If a request is put from the same task that allocated it,
then we can maintain a per-ctx cache of them. This obviously relies
on io_uring always pruning caches in a reliable way, and there's
currently a case off io_uring fd release where we can miss that.
One example is a ring setup with IOPOLL, which relies on the task
polling for completions, which will free them. However, if such a task
submits a request and then exits or closes the ring without reaping
the completion, then ring release will reap and put. If release happens
from that very same task, the completed request task refs will get
put back into the cache pool. This is problematic, as we're now beyond
the point of pruning caches.
Manually drop these caches after doing an IOPOLL reap. This releases
references from the current task, which is enough. If another task
happens to be doing the release, then the caching will not be
triggered and there's no issue.
Cc: stable@vger.kernel.org
Fixes: e98e49b2bb ("io_uring: extend task put optimisations")
Reported-by: Homin Rhee <hominlab@gmail.com>
Change-Id: I9495121af065424141fa9c39840ab9aa91f45c72
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Bug: 268174392
(cherry picked from commit e9c6556708)
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
