Oliver Neukum [Thu, 8 Aug 2019 09:27:28 +0000 (11:27 +0200)]
usb: iowarrior: fix deadlock on disconnect
We have to drop the mutex before we close() upon disconnect()
as close() needs the lock. This is safe to do by dropping the
mutex as intfdata is already set to NULL, so open() will fail.
Colin Ian King [Mon, 24 Jun 2019 16:39:59 +0000 (09:39 -0700)]
drm/vmwgfx: fix memory leak when too many retries have occurred
Currently when too many retries have occurred there is a memory
leak on the allocation for reply on the error return path. Fix
this by kfree'ing reply before returning.
Addresses-Coverity: ("Resource leak") Fixes: a9cd9c044aa9 ("drm/vmwgfx: Add a check to handle host message failure") Signed-off-by: Colin Ian King <[email protected]> Reviewed-by: Deepak Rawat <[email protected]> Signed-off-by: Deepak Rawat <[email protected]> Signed-off-by: Thomas Hellstrom <[email protected]>
Wenwen Wang [Thu, 8 Aug 2019 05:50:58 +0000 (00:50 -0500)]
ALSA: firewire: fix a memory leak bug
In iso_packets_buffer_init(), 'b->packets' is allocated through
kmalloc_array(). Then, the aligned packet size is checked. If it is
larger than PAGE_SIZE, -EINVAL will be returned to indicate the error.
However, the allocated 'b->packets' is not deallocated on this path,
leading to a memory leak.
To fix the above issue, free 'b->packets' before returning the error code.
Ming Lei [Mon, 5 Aug 2019 01:19:06 +0000 (09:19 +0800)]
genirq/affinity: Create affinity mask for single vector
Since commit c66d4bd110a1f8 ("genirq/affinity: Add new callback for
(re)calculating interrupt sets"), irq_create_affinity_masks() returns
NULL in case of single vector. This change has caused regression on some
drivers, such as lpfc.
The problem is that single vector requests can happen in some generic cases:
1) kdump kernel
2) irq vectors resource is close to exhaustion.
If in that situation the affinity mask for a single vector is not created,
every caller has to handle the special case.
There is no reason why the mask cannot be created, so remove the check for
a single vector and create the mask.
Tony writes:
Somehow this causes a regression in Linux next for me where I'm
seeing lots of sysfs entries now missing under
/sys/bus/platform/devices.
For example, I now only see one .serial entry show up in sysfs.
Things work again if I revert commit cc798c83898e ("kernfs: fix
memleak inkernel_ops_readdir()"). Any ideas why that would be?
Tejun says:
Ugh, you're right. It can get double-put cuz ctx->pos is put by
release too.
When building with W=1, warnings about missing prototypes are emitted:
CC arch/x86/lib/cpu.o
arch/x86/lib/cpu.c:5:14: warning: no previous prototype for 'x86_family' [-Wmissing-prototypes]
5 | unsigned int x86_family(unsigned int sig)
| ^~~~~~~~~~
arch/x86/lib/cpu.c:18:14: warning: no previous prototype for 'x86_model' [-Wmissing-prototypes]
18 | unsigned int x86_model(unsigned int sig)
| ^~~~~~~~~
arch/x86/lib/cpu.c:33:14: warning: no previous prototype for 'x86_stepping' [-Wmissing-prototypes]
33 | unsigned int x86_stepping(unsigned int sig)
| ^~~~~~~~~~~~
Add the proper include file so the prototypes are there.
x86/purgatory: Use CFLAGS_REMOVE rather than reset KBUILD_CFLAGS
KBUILD_CFLAGS is very carefully built up in the top level Makefile,
particularly when cross compiling or using different build tools.
Resetting KBUILD_CFLAGS via := assignment is an antipattern.
The comment above the reset mentions that -pg is problematic. Other
Makefiles use `CFLAGS_REMOVE_file.o = $(CC_FLAGS_FTRACE)` when
CONFIG_FUNCTION_TRACER is set. Prefer that pattern to wiping out all of
the important KBUILD_CFLAGS then manually having to re-add them. Seems
also that __stack_chk_fail references are generated when using
CONFIG_STACKPROTECTOR or CONFIG_STACKPROTECTOR_STRONG.
x86/purgatory: Do not use __builtin_memcpy and __builtin_memset
Implementing memcpy and memset in terms of __builtin_memcpy and
__builtin_memset is problematic.
GCC at -O2 will replace calls to the builtins with calls to memcpy and
memset (but will generate an inline implementation at -Os). Clang will
replace the builtins with these calls regardless of optimization level.
$ llvm-objdump -dr arch/x86/purgatory/string.o | tail
Such code results in infinite recursion at runtime. This is observed
when doing kexec.
Instead, reuse an implementation from arch/x86/boot/compressed/string.c.
This requires to implement a stub function for warn(). Also, Clang may
lower memcmp's that compare against 0 to bcmp's, so add a small definition,
too. See also: commit 5f074f3e192f ("lib/string.c: implement a basic bcmp")
Wenwen Wang [Thu, 8 Aug 2019 05:15:21 +0000 (00:15 -0500)]
sound: fix a memory leak bug
In sound_insert_unit(), the controlling structure 's' is allocated through
kmalloc(). Then it is added to the sound driver list by invoking
__sound_insert_unit(). Later on, if __register_chrdev() fails, 's' is
removed from the list through __sound_remove_unit(). If 'index' is not less
than 0, -EBUSY is returned to indicate the error. However, 's' is not
deallocated on this execution path, leading to a memory leak bug.
To fix the above issue, free 's' before -EBUSY is returned.
Dave Airlie [Thu, 8 Aug 2019 03:25:49 +0000 (13:25 +1000)]
Merge tag 'drm-fixes-5.3-2019-08-07' of git://people.freedesktop.org/~agd5f/linux into drm-fixes
drm-fixes-5.3-2019-08-07:
amdgpu:
- Fixes VCN to handle the latest navi10 firmware
- Fixes for fan control on navi10
- Properly handle SMU metrics table on navi10
- Fix a resume regression on Stoney
James Smart [Fri, 2 Aug 2019 20:26:12 +0000 (13:26 -0700)]
scsi: lpfc: Fix crash when cpu count is 1 and null irq affinity mask
When a configurations runs with a single cpu (such as a kdump kernel),
which causes the driver to request a single vector, when the driver
subsequently requests an irq affinity mask, the mask comes back null. The
driver currently does nothing in this scenario, which leaves mappings to
hardware queues incomplete and crashes the system.
Fix by recognizing the null mask and assigning the vector to the first cpu
in the system.
hwmon: (nct7802) Fix wrong detection of in4 presence
The code to detect if in4 is present is wrong; if in4 is not present,
the in4_input sysfs attribute is still present.
In detail:
- Ihen RTD3_MD=11 (VSEN3 present), everything is as expected (no bug).
- If we have RTD3_MD!=11 (no VSEN3), we unexpectedly have a in4_input
file under /sys and the "sensors" command displays in4_input.
But as expected, we have no in4_min, in4_max, in4_alarm, in4_beep.
Fix is_visible function to detect and report in4_input visibility
as expected.
Yishai Hadas [Mon, 5 Aug 2019 08:30:10 +0000 (11:30 +0300)]
IB/mlx5: Fix implicit MR release flow
Once implicit MR is being called to be released by
ib_umem_notifier_release() its leaves were marked as "dying".
However, when dereg_mr()->mlx5_ib_free_implicit_mr()->mr_leaf_free() is
called, it skips running the mr_leaf_free_action (i.e. umem_odp->work)
when those leaves were marked as "dying".
As such ib_umem_release() for the leaves won't be called and their MRs
will be leaked as well.
When an application exits/killed without calling dereg_mr we might hit the
above flow.
This fatal scenario is reported by WARN_ON() upon
mlx5_ib_dealloc_ucontext() as ibcontext->per_mm_list is not empty, the
call trace can be seen below.
Originally the "dying" mark as part of ib_umem_notifier_release() was
introduced to prevent pagefault_mr() from returning a success response
once this happened. However, we already have today the completion
mechanism so no need for that in those flows any more. Even in case a
success response will be returned the firmware will not find the pages and
an error will be returned in the following call as a released mm will
cause ib_umem_odp_map_dma_pages() to permanently fail mmget_not_zero().
Fix the above issue by dropping the "dying" from the above flows. The
other flows that are using "dying" are still needed it for their
synchronization purposes.
Jens Axboe [Wed, 7 Aug 2019 18:23:57 +0000 (12:23 -0600)]
libata: add SG safety checks in SFF pio transfers
Abort processing of a command if we run out of mapped data in the
SG list. This should never happen, but a previous bug caused it to
be possible. Play it safe and attempt to abort nicely if we don't
have more SG segments left.
Jens Axboe [Wed, 7 Aug 2019 18:20:52 +0000 (12:20 -0600)]
libata: have ata_scsi_rw_xlat() fail invalid passthrough requests
For passthrough requests, libata-scsi takes what the user passes in
as gospel. This can be problematic if the user fills in the CDB
incorrectly. One example of that is in request sizes. For read/write
commands, the CDB contains fields describing the transfer length of
the request. These should match with the SG_IO header fields, but
libata-scsi currently does no validation of that.
Check that the number of blocks in the CDB for passthrough requests
matches what was mapped into the request. If the CDB asks for more
data then the validated SG_IO header fields, error it.
Jens Axboe [Tue, 6 Aug 2019 19:34:31 +0000 (13:34 -0600)]
block: fix O_DIRECT error handling for bio fragments
0eb6ddfb865c tried to fix this up, but introduced a use-after-free
of dio. Additionally, we still had an issue with error handling,
as reported by Darrick:
"I noticed a regression in xfs/747 (an unreleased xfstest for the
xfs_scrub media scanning feature) on 5.3-rc3. I'll condense that down
to a simpler reproducer:
error-test: 0 209 linear 8:48 0
error-test: 209 1 error
error-test: 210 6446894 linear 8:48 210
Basically we have a ~3G /dev/sdd and we set up device mapper to fail IO
for sector 209 and to pass the io to the scsi device everywhere else.
On 5.3-rc3, performing a directio pread of this range with a < 1M buffer
(in other words, a request for fewer than MAX_BIO_PAGES bytes) yields
EIO like you'd expect:
pread64(3, "XFSB\0\0\20\0\0\0\0\0\0\fL\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 1146880, 0) = 1146880
read 1146880/1146880 bytes at offset 0
1 MiB, 1 ops; 0.0009 sec (1.124 GiB/sec and 1052.6316 ops/sec)
+++ exited with 0 +++
(Note that the part of the buffer corresponding to the dm-error area is
uninitialized)
On 5.3-rc2, both commands would fail with EIO like you'd expect. The
only change between rc2 and rc3 is commit 0eb6ddfb865c ("block: Fix
__blkdev_direct_IO() for bio fragments").
AFAICT we end up in __blkdev_direct_IO with a 1120K buffer, which gets
split into two bios: one for the first BIO_MAX_PAGES worth of data (1MB)
and a second one for the 96k after that."
Fix this by noting that it's always safe to dereference dio if we get
BLK_QC_T_EAGAIN returned, as end_io hasn't been run for that case. So
we can safely increment the dio size before calling submit_bio(), and
then decrement it on failure (not that it really matters, as the bio
and dio are going away).
For error handling, return to the original method of just using 'ret'
for tracking the error, and the size tracking in dio->size.
Fixes: 0eb6ddfb865c ("block: Fix __blkdev_direct_IO() for bio fragments") Fixes: 6a43074e2f46 ("block: properly handle IOCB_NOWAIT for async O_DIRECT IO") Reported-by: Darrick J. Wong <[email protected]> Signed-off-by: Jens Axboe <[email protected]>
Make HMM_MIRROR an option that is selected by drivers wanting to use it
instead of a user visible option as it is just a low-level implementation
detail.
mm/hmm: cleanup the hmm_vma_walk_hugetlb_entry stub
Stub out the whole function and assign NULL to the .hugetlb_entry method
if CONFIG_HUGETLB_PAGE is not set, as the method won't ever be called in
that case.
We only need the special pud_entry walker if PUD-sized hugepages and pte
mappings are supported, else the common pagewalk code will take care of
the iteration. Not implementing this callback reduced the amount of code
compiled for non-x86 platforms, and also fixes compile failures with other
architectures when helpers like pud_pfn are not implemented.
mm/hmm: don't abuse pte_index() in hmm_vma_handle_pmd
pte_index is an internal arch helper in various architectures, without
consistent semantics. Open code that calculation of a PMD index based on
the virtual address instead.
mm/hmm: remove the page_shift member from struct hmm_range
All users pass PAGE_SIZE here, and if we wanted to support single entries
for huge pages we should really just add a HMM_FAULT_HUGEPAGE flag instead
that uses the huge page size instead of having the caller calculate that
size once, just for the hmm code to verify it.
nouveau: pass struct nouveau_svmm to nouveau_range_fault
We'll need the nouveau_svmm structure to improve the function soon. For
now this allows using the svmm->mm reference to unlock the mmap_sem, and
thus the same dereference chain that the caller uses to lock and unlock
it.
amdgpu: don't initialize range->list in amdgpu_hmm_init_range
The list is used to add the range to another list as an entry in the core
hmm code, and intended as a private member not exposed to drivers. There
is no need to initialize it in a driver.
amdgpu: remove -EAGAIN handling for hmm_range_fault
hmm_range_fault can only return -EAGAIN if called with the
HMM_FAULT_ALLOW_RETRY flag, which amdgpu never does. Remove the handling
for the -EAGAIN case with its non-standard locking scheme.
Filipe Manana [Wed, 7 Aug 2019 10:21:46 +0000 (11:21 +0100)]
Btrfs: fix sysfs warning and missing raid sysfs directories
In the 5.3 merge window, commit 7c7e301406d0a9 ("btrfs: sysfs: Replace
default_attrs in ktypes with groups"), we started using the member
"defaults_groups" for the kobject type "btrfs_raid_ktype". That leads
to a series of warnings when running some test cases of fstests, such
as btrfs/027, btrfs/124 and btrfs/176. The traces produced by those
warnings are like the following:
These happen because, during the unmount path, we call kobject_del() for
raid kobjects that are not fully initialized, meaning that we set their
ktype (as btrfs_raid_ktype) through link_block_group() but we didn't set
their parent kobject, which is done through btrfs_add_raid_kobjects().
We have this split raid kobject setup since commit 75cb379d263521
("btrfs: defer adding raid type kobject until after chunk relocation") in
order to avoid triggering reclaim during contextes where we can not
(either we are holding a transaction handle or some lock required by
the transaction commit path), so that we do the calls to kobject_add(),
which triggers GFP_KERNEL allocations, through btrfs_add_raid_kobjects()
in contextes where it is safe to trigger reclaim. That change expected
that a new raid kobject can only be created either when mounting the
filesystem or after raid profile conversion through the relocation path.
However, we can have new raid kobject created in other two cases at least:
1) During device replace (or scrub) after adding a device a to the
filesystem. The replace procedure (and scrub) do calls to
btrfs_inc_block_group_ro() which can allocate a new block group
with a new raid profile (because we now have more devices). This
can be triggered by test cases btrfs/027 and btrfs/176.
2) During a degraded mount trough any write path. This can be triggered
by test case btrfs/124.
Fixing this by adding extra calls to btrfs_add_raid_kobjects(), not only
makes things more complex and fragile, can also introduce deadlocks with
reclaim the following way:
1) Calling btrfs_add_raid_kobjects() at btrfs_inc_block_group_ro() or
anywhere in the replace/scrub path will cause a deadlock with reclaim
because if reclaim happens and a transaction commit is triggered,
the transaction commit path will block at btrfs_scrub_pause().
2) During degraded mounts it is essentially impossible to figure out where
to add extra calls to btrfs_add_raid_kobjects(), because allocation of
a block group with a new raid profile can happen anywhere, which means
we can't safely figure out which contextes are safe for reclaim, as
we can either hold a transaction handle or some lock needed by the
transaction commit path.
So it is too complex and error prone to have this split setup of raid
kobjects. So fix the issue by consolidating the setup of the kobjects in a
single place, at link_block_group(), and setup a nofs context there in
order to prevent reclaim being triggered by the memory allocations done
through the call chain of kobject_add().
Besides fixing the sysfs warnings during kobject_del(), this also ensures
the sysfs directories for the new raid profiles end up created and visible
to users (a bug that existed before the 5.3 commit 7c7e301406d0a9
("btrfs: sysfs: Replace default_attrs in ktypes with groups")).
Fixes: 75cb379d263521 ("btrfs: defer adding raid type kobject until after chunk relocation") Fixes: 7c7e301406d0a9 ("btrfs: sysfs: Replace default_attrs in ktypes with groups") Signed-off-by: Filipe Manana <[email protected]> Reviewed-by: David Sterba <[email protected]> Signed-off-by: David Sterba <[email protected]>
Mark switch cases where we are expecting to fall through.
Fix the following warning (Building: allnoconfig i386):
arch/x86/kernel/ptrace.c:202:6: warning: this statement may fall through [-Wimplicit-fallthrough=]
if (unlikely(value == 0))
^
arch/x86/kernel/ptrace.c:206:2: note: here
default:
^~~~~~~
Takashi Iwai [Tue, 6 Aug 2019 15:31:48 +0000 (17:31 +0200)]
ALSA: hda - Workaround for crackled sound on AMD controller (1022:1457)
A long-time problem on the recent AMD chip (X370, X470, B450, etc with
PCI ID 1022:1457) with Realtek codecs is the crackled or distorted
sound for capture streams, as well as occasional playback hiccups.
After lengthy debugging sessions, the workarounds we've found are like
the following:
- Set up the proper driver caps for this controller, similar as the
other AMD controller.
- Correct the DMA position reporting with the fixed FIFO size, which
is similar like as workaround used for VIA chip set.
- Even after the position correction, PulseAudio still shows
mysterious stalls of playback streams when a capture is triggered in
timer-scheduled mode. Since we have no clear way to eliminate the
stall, pass the BATCH PCM flag for PA to suppress the tsched mode as
a temporary workaround.
This patch implements the workarounds. For the driver caps, it
defines a new preset, AXZ_DCAPS_PRESET_AMD_SB. It enables the FIFO-
corrected position reporting (corresponding to the new position_fix=6)
and enforces the SNDRV_PCM_INFO_BATCH flag.
Note that the current implementation is merely a workaround.
Hopefully we'll find a better alternative in future, especially about
removing the BATCH flag hack again.
Mika Westerberg [Wed, 7 Aug 2019 10:57:18 +0000 (13:57 +0300)]
Revert "PCI: Add missing link delays required by the PCIe spec"
Commit c2bf1fc212f7 ("PCI: Add missing link delays required by the PCIe
spec") turned out causing issues with some systems either by making them
unresponsive or slowing down runtime and system wide resume of PCIe
devices. While root cause for the unresponsiveness is still under
investigation given the amount of issues reported better to revert it
for now.
Wenwen Wang [Wed, 7 Aug 2019 09:08:51 +0000 (04:08 -0500)]
ALSA: hiface: fix multiple memory leak bugs
In hiface_pcm_init(), 'rt' is firstly allocated through kzalloc(). Later
on, hiface_pcm_init_urb() is invoked to initialize 'rt->out_urbs[i]'. In
hiface_pcm_init_urb(), 'rt->out_urbs[i].buffer' is allocated through
kzalloc(). However, if hiface_pcm_init_urb() fails, both 'rt' and
'rt->out_urbs[i].buffer' are not deallocated, leading to memory leak bugs.
Also, 'rt->out_urbs[i].buffer' is not deallocated if snd_pcm_new() fails.
To fix the above issues, free 'rt' and 'rt->out_urbs[i].buffer'.
Pull networking fixes from David Miller:
"Yeah I should have sent a pull request last week, so there is a lot
more here than usual:
1) Fix memory leak in ebtables compat code, from Wenwen Wang.
2) Several kTLS bug fixes from Jakub Kicinski (circular close on
disconnect etc.)
3) Force slave speed check on link state recovery in bonding 802.3ad
mode, from Thomas Falcon.
4) Clear RX descriptor bits before assigning buffers to them in
stmmac, from Jose Abreu.
5) Several missing of_node_put() calls, mostly wrt. for_each_*() OF
loops, from Nishka Dasgupta.
6) Double kfree_skb() in peak_usb can driver, from Stephane Grosjean.
7) Need to hold sock across skb->destructor invocation, from Cong
Wang.
8) IP header length needs to be validated in ipip tunnel xmit, from
Haishuang Yan.
9) Use after free in ip6 tunnel driver, also from Haishuang Yan.
10) Do not use MSI interrupts on r8169 chips before RTL8168d, from
Heiner Kallweit.
11) Upon bridge device init failure, we need to delete the local fdb.
From Nikolay Aleksandrov.
12) Handle erros from of_get_mac_address() properly in stmmac, from
Martin Blumenstingl.
13) Handle concurrent rename vs. dump in netfilter ipset, from Jozsef
Kadlecsik.
14) Setting NETIF_F_LLTX on mac80211 causes complete breakage with
some devices, so revert. From Johannes Berg.
15) Fix deadlock in rxrpc, from David Howells.
16) Fix Kconfig deps of enetc driver, we must have PHYLIB. From Yue
Haibing.
17) Fix mvpp2 crash on module removal, from Matteo Croce.
18) Fix race in genphy_update_link, from Heiner Kallweit.
19) bpf_xdp_adjust_head() stopped working with generic XDP when we
fixes generic XDP to support stacked devices properly, fix from
Jesper Dangaard Brouer.
20) Unbalanced RCU locking in rt6_update_exception_stamp_rt(), from
David Ahern.
21) Several memory leaks in new sja1105 driver, from Vladimir Oltean"
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (214 commits)
net: dsa: sja1105: Fix memory leak on meta state machine error path
net: dsa: sja1105: Fix memory leak on meta state machine normal path
net: dsa: sja1105: Really fix panic on unregistering PTP clock
net: dsa: sja1105: Use the LOCKEDS bit for SJA1105 E/T as well
net: dsa: sja1105: Fix broken learning with vlan_filtering disabled
net: dsa: qca8k: Add of_node_put() in qca8k_setup_mdio_bus()
net: sched: sample: allow accessing psample_group with rtnl
net: sched: police: allow accessing police->params with rtnl
net: hisilicon: Fix dma_map_single failed on arm64
net: hisilicon: fix hip04-xmit never return TX_BUSY
net: hisilicon: make hip04_tx_reclaim non-reentrant
tc-testing: updated vlan action tests with batch create/delete
net sched: update vlan action for batched events operations
net: stmmac: tc: Do not return a fragment entry
net: stmmac: Fix issues when number of Queues >= 4
net: stmmac: xgmac: Fix XGMAC selftests
be2net: disable bh with spin_lock in be_process_mcc
net: cxgb3_main: Fix a resource leak in a error path in 'init_one()'
net: ethernet: sun4i-emac: Support phy-handle property for finding PHYs
net: bridge: move default pvid init/deinit to NETDEV_REGISTER/UNREGISTER
...
David S. Miller [Tue, 6 Aug 2019 21:37:02 +0000 (14:37 -0700)]
Merge branch 'sja1105-fixes'
Vladimir Oltean says:
====================
Fixes for SJA1105 DSA: FDBs, Learning and PTP
This is an assortment of functional fixes for the sja1105 switch driver
targeted for the "net" tree (although they apply on net-next just as
well).
Patch 1/5 ("net: dsa: sja1105: Fix broken learning with vlan_filtering
disabled") repairs a breakage introduced in the early development stages
of the driver: support for traffic from the CPU has broken "normal"
frame forwarding (based on DMAC) - there is connectivity through the
switch only because all frames are flooded.
I debated whether this patch qualifies as a fix, since it puts the
switch into a mode it has never operated in before (aka SVL). But
"normal" forwarding did use to work before the "Traffic support for
SJA1105 DSA driver" patchset, and arguably this patch should have been
part of that.
Also, it would be strange for this feature to be broken in the 5.2 LTS.
Patch 2/5 ("net: dsa: sja1105: Use the LOCKEDS bit for SJA1105 E/T as
well") is a simplification of a previous FDB-related patch that is
currently in the 5.3 rc's.
Patches 3/5 - 5/5 fix various crashes found while running linuxptp over the
switch ports for extended periods of time, or in conjunction with other
error conditions. The fixed-up commits were all introduced in 5.2.
====================
Vladimir Oltean [Sun, 4 Aug 2019 22:38:48 +0000 (01:38 +0300)]
net: dsa: sja1105: Fix memory leak on meta state machine error path
When RX timestamping is enabled and two link-local (non-meta) frames are
received in a row, this constitutes an error.
The tagger is always caching the last link-local frame, in an attempt to
merge it with the meta follow-up frame when that arrives. To recover
from the above error condition, the initial cached link-local frame is
dropped and the second frame in a row is cached (in expectance of the
second meta frame).
However, when dropping the initial link-local frame, its backing memory
was being leaked.
Fixes: f3097be21bf1 ("net: dsa: sja1105: Add a state machine for RX timestamping") Signed-off-by: Vladimir Oltean <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Vladimir Oltean [Sun, 4 Aug 2019 22:38:47 +0000 (01:38 +0300)]
net: dsa: sja1105: Fix memory leak on meta state machine normal path
After a meta frame is received, it is associated with the cached
sp->data->stampable_skb from the DSA tagger private structure.
Cached means its refcount is incremented with skb_get() in order for
dsa_switch_rcv() to not free it when the tagger .rcv returns NULL.
The mistake is that skb_unref() is not the correct function to use. It
will correctly decrement the refcount (which will go back to zero) but
the skb memory will not be freed. That is the job of kfree_skb(), which
also calls skb_unref().
But it turns out that freeing the cached stampable_skb is in fact not
necessary. It is still a perfectly valid skb, and now it is even
annotated with the partial RX timestamp. So remove the skb_copy()
altogether and simply pass the stampable_skb with a refcount of 1
(incremented by us, decremented by dsa_switch_rcv) up the stack.
Fixes: f3097be21bf1 ("net: dsa: sja1105: Add a state machine for RX timestamping") Signed-off-by: Vladimir Oltean <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Vladimir Oltean [Sun, 4 Aug 2019 22:38:46 +0000 (01:38 +0300)]
net: dsa: sja1105: Really fix panic on unregistering PTP clock
The IS_ERR_OR_NULL(priv->clock) check inside
sja1105_ptp_clock_unregister() is preventing cancel_delayed_work_sync
from actually being run.
Additionally, sja1105_ptp_clock_unregister() does not actually get run,
when placed in sja1105_remove(). The DSA switch gets torn down, but the
sja1105 module does not get unregistered. So sja1105_ptp_clock_unregister
needs to be moved to sja1105_teardown, to be symmetrical with
sja1105_ptp_clock_register which is called from the DSA sja1105_setup.
It is strange to fix a "fixes" patch, but the probe failure can only be
seen when the attached PHY does not respond to MDIO (issue which I can't
pinpoint the reason to) and it goes away after I power-cycle the board.
This time the patch was validated on a failing board, and the kernel
panic from the fixed commit's message can no longer be seen.
Fixes: 29dd908d355f ("net: dsa: sja1105: Cancel PTP delayed work on unregister") Signed-off-by: Vladimir Oltean <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Vladimir Oltean [Sun, 4 Aug 2019 22:38:45 +0000 (01:38 +0300)]
net: dsa: sja1105: Use the LOCKEDS bit for SJA1105 E/T as well
It looks like the FDB dump taken from first-generation switches also
contains information on whether entries are static or not. So use that
instead of searching through the driver's tables.
Fixes: d763778224ea ("net: dsa: sja1105: Implement is_static for FDB entries on E/T") Signed-off-by: Vladimir Oltean <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Vladimir Oltean [Sun, 4 Aug 2019 22:38:44 +0000 (01:38 +0300)]
net: dsa: sja1105: Fix broken learning with vlan_filtering disabled
When put under a bridge with vlan_filtering 0, the SJA1105 ports will
flood all traffic as if learning was broken. This is because learning
interferes with the rx_vid's configured by dsa_8021q as unique pvid's.
So learning technically still *does* work, it's just that the learnt
entries never get matched due to their unique VLAN ID.
The setting that saves the day is Shared VLAN Learning, which on this
switch family works exactly as desired: VLAN tagging still works
(untagged traffic gets the correct pvid) and FDB entries are still
populated with the correct contents including VID. Also, a frame cannot
violate the forwarding domain restrictions enforced by its classified
VLAN. It is just that the VID is ignored when looking up the FDB for
taking a forwarding decision (selecting the egress port).
This patch activates SVL, and the result is that frames with a learnt
DMAC are no longer flooded in the scenario described above.
Now exactly *because* SVL works as desired, we have to revisit some
earlier patches:
- It is no longer necessary to manipulate the VID of the 'bridge fdb
{add,del}' command when vlan_filtering is off. This is because now,
SVL is enabled for that case, so the actual VID does not matter*.
- It is still desirable to hide dsa_8021q VID's in the FDB dump
callback. But right now the dump callback should no longer hide
duplicates (one per each front panel port's pvid, plus one for the
VLAN that the CPU port is going to tag a TX frame with), because there
shouldn't be any (the switch will match a single FDB entry no matter
its VID anyway).
* Not really... It's no longer necessary to transform a 'bridge fdb add'
into 5 fdb add operations, but the user might still add a fdb entry with
any vid, and all of them would appear as duplicates in 'bridge fdb
show'. So force a 'bridge fdb add' to insert the VID of 0**, so that we
can prune the duplicates at insertion time.
** The VID of 0 is better than 1 because it is always guaranteed to be
in the ports' hardware filter. DSA also avoids putting the VID inside
the netlink response message towards the bridge driver when we return
this particular VID, which makes it suitable for FDB entries learnt
with vlan_filtering off.
Fixes: 227d07a07ef1 ("net: dsa: sja1105: Add support for traffic through standalone ports") Signed-off-by: Vladimir Oltean <[email protected]> Signed-off-by: Georg Waibel <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Nishka Dasgupta [Sun, 4 Aug 2019 15:30:18 +0000 (21:00 +0530)]
net: dsa: qca8k: Add of_node_put() in qca8k_setup_mdio_bus()
Each iteration of for_each_available_child_of_node() puts the previous
node, but in the case of a return from the middle of the loop, there
is no put, thus causing a memory leak. Hence add an of_node_put() before
the return.
Additionally, the local variable ports in the function
qca8k_setup_mdio_bus() takes the return value of of_get_child_by_name(),
which gets a node but does not put it. If the function returns without
putting ports, it may cause a memory leak. Hence put ports before the
mid-loop return statement, and also outside the loop after its last usage
in this function.
Issues found with Coccinelle.
David S. Miller [Tue, 6 Aug 2019 21:15:39 +0000 (14:15 -0700)]
Merge branch 'flow_offload-action-fixes'
Vlad Buslov says:
====================
action fixes for flow_offload infra compatibility
Fix rcu warnings due to usage of action helpers that expect rcu read lock
protection from rtnl-protected context of flow_offload infra.
====================
Change tcf_sample_psample_group() helper to allow using it from both rtnl
and rcu protected contexts.
Fixes: a7a7be6087b0 ("net/sched: add sample action to the hardware intermediate representation") Signed-off-by: Vlad Buslov <[email protected]> Reviewed-by: Pieter Jansen van Vuuren <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Change tcf_police_rate_bytes_ps() and tcf_police_tcfp_burst() helpers to
allow using them from both rtnl and rcu protected contexts.
Fixes: 8c8cfc6ed274 ("net/sched: add police action to the hardware intermediate representation") Signed-off-by: Vlad Buslov <[email protected]> Reviewed-by: Pieter Jansen van Vuuren <[email protected]> Signed-off-by: David S. Miller <[email protected]>
David S. Miller [Tue, 6 Aug 2019 21:14:01 +0000 (14:14 -0700)]
Merge branch 'hisilicon-fixes'
Jiangfeng Xiao says:
====================
net: hisilicon: Fix a few problems with hip04_eth
During the use of the hip04_eth driver,
several problems were found,
which solved the hip04_tx_reclaim reentry problem,
fixed the problem that hip04_mac_start_xmit never
returns NETDEV_TX_BUSY
and the dma_map_single failed on the arm64 platform.
====================
Jiangfeng Xiao [Sat, 3 Aug 2019 12:31:41 +0000 (20:31 +0800)]
net: hisilicon: Fix dma_map_single failed on arm64
On the arm64 platform, executing "ifconfig eth0 up" will fail,
returning "ifconfig: SIOCSIFFLAGS: Input/output error."
ndev->dev is not initialized, dma_map_single->get_dma_ops->
dummy_dma_ops->__dummy_map_page will return DMA_ERROR_CODE
directly, so when we use dma_map_single, the first parameter
is to use the device of platform_device.
Jiangfeng Xiao [Sat, 3 Aug 2019 12:31:40 +0000 (20:31 +0800)]
net: hisilicon: fix hip04-xmit never return TX_BUSY
TX_DESC_NUM is 256, in tx_count, the maximum value of
mod(TX_DESC_NUM - 1) is 254, the variable "count" in
the hip04_mac_start_xmit function is never equal to
(TX_DESC_NUM - 1), so hip04_mac_start_xmit never
return NETDEV_TX_BUSY.
tx_count is modified to mod(TX_DESC_NUM) so that
the maximum value of tx_count can reach
(TX_DESC_NUM - 1), then hip04_mac_start_xmit can reurn
NETDEV_TX_BUSY.
Jiangfeng Xiao [Sat, 3 Aug 2019 12:31:39 +0000 (20:31 +0800)]
net: hisilicon: make hip04_tx_reclaim non-reentrant
If hip04_tx_reclaim is interrupted while it is running
and then __napi_schedule continues to execute
hip04_rx_poll->hip04_tx_reclaim, reentrancy occurs
and oops is generated. So you need to mask the interrupt
during the hip04_tx_reclaim run.
====================
Fix batched event generation for vlan action
When adding or deleting a batch of entries, the kernel sends up to
TCA_ACT_MAX_PRIO (defined to 32 in kernel) entries in an event to user
space. However it does not consider that the action sizes may vary and
require different skb sizes.
For example, consider the following script adding 32 entries with all
supported vlan parameters (in order to maximize netlink messages size):
$TC actions flush action vlan
for i in `seq 1 $1`;
do
cmd="action vlan push protocol 802.1q id 4094 priority 7 pipe \
index $i cookie aabbccddeeff112233445566778800a1 "
args=$args$cmd
done
$TC actions add $args
%
% ./tc-batch.sh 32
Error: Failed to fill netlink attributes while adding TC action.
We have an error talking to the kernel
%
patch 1 adds callback in tc_action_ops of vlan action, which calculates
the action size, and passes size to tcf_add_notify()/tcf_del_notify().
patch 2 updates the TDC test suite with relevant vlan test cases.
====================
Linus Torvalds [Tue, 6 Aug 2019 21:01:08 +0000 (14:01 -0700)]
Merge tag 'mips_fixes_5.3_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux
Pull MIPS fixes from Paul Burton:
"A few MIPS fixes for 5.3:
- Various switch fall through annotations to fixup warnings & errors
resulting from -Wimplicit-fallthrough.
- A fix for systems (at least jazz) using an i8253 PIT as clocksource
when it's not suitably configured.
- Set struct cacheinfo's cpu_map_populated field to true, indicating
that we filled in cache info detected from cop0 registers &
avoiding complaints about that info being (intentionally) missing
in devicetree"
* tag 'mips_fixes_5.3_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
MIPS: BCM63XX: Mark expected switch fall-through
MIPS: OProfile: Mark expected switch fall-throughs
MIPS: Annotate fall-through in Cavium Octeon code
MIPS: Annotate fall-through in kvm/emulate.c
mips: fix cacheinfo
MIPS: kernel: only use i8253 clocksource with periodic clockevent
Likun Gao [Fri, 2 Aug 2019 07:18:57 +0000 (15:18 +0800)]
drm/amdgpu: pin the csb buffer on hw init for gfx v8
Without this pin, the csb buffer will be filled with inconsistent
data after S3 resume. And that will causes gfx hang on gfxoff
exit since this csb will be executed then.
Jose Abreu [Tue, 6 Aug 2019 13:16:16 +0000 (15:16 +0200)]
net: stmmac: xgmac: Fix XGMAC selftests
Fixup the XGMAC selftests by correctly finishing the implementation of
set_filter callback.
Result:
$ ethtool -t enp4s0
The test result is PASS
The test extra info:
1. MAC Loopback 0
2. PHY Loopback -95
3. MMC Counters -95
4. EEE -95
5. Hash Filter MC 0
6. Perfect Filter UC 0
7. MC Filter 0
8. UC Filter 0
9. Flow Control 0
David S. Miller [Tue, 6 Aug 2019 18:48:48 +0000 (11:48 -0700)]
Merge tag 'wireless-drivers-for-davem-2019-08-06' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers
Kalle Valo says:
====================
wireless-drivers fixes for 5.3
Second set of fixes for 5.3. Lots of iwlwifi fixes have accumulated
which consists most of patches in this pull request. Only most notable
iwlwifi fixes are listed below.
mwifiex
* fix a regression related to WPA1 networks since v5.3-rc1
iwlwifi
* fix use-after-free issues
* fix DMA mapping API usage errors
* fix frame drop occurring due to reorder buffer handling in
RSS in certain conditions
* fix rate scale locking issues
* disable TX A-MSDU on older NICs as it causes problems and was
never supposed to be supported
* new PCI IDs
* GEO_TX_POWER_LIMIT API issue that many people were hitting
====================
Linus Torvalds [Tue, 6 Aug 2019 18:47:23 +0000 (11:47 -0700)]
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid
Pull HID fixes from Jiri Kosina:
- functional regression fix for some of the Logitech unifying devices,
from Hans de Goede
- race condition fix in hid-sony for bug severely affecting
Valve/Android deployments, from Roderick Colenbrander
- several fixes for issues found by syzbot/kasan, from Oliver Neukum
and Hillf Danton
- functional regression fix for Wacom Cintiq device, from Aaron
Armstrong Skomra
- a few other assorted device-specific quirks
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid:
HID: sony: Fix race condition between rumble and device remove.
HID: hiddev: do cleanup in failure of opening a device
HID: hiddev: avoid opening a disconnected device
HID: input: fix a4tech horizontal wheel custom usage
HID: Add quirk for HP X1200 PIXART OEM mouse
HID: holtek: test for sanity of intfdata
HID: wacom: fix bit shift for Cintiq Companion 2
HID: quirks: Set the INCREMENT_USAGE_ON_DUPLICATE quirk on Saitek X52
HID: logitech-dj: Really fix return value of logi_dj_recv_query_hidpp_devices
HID: Add 044f:b320 ThrustMaster, Inc. 2 in 1 DT
HID: logitech-dj: add the Powerplay receiver
HID: logitech-hidpp: add USB PID for a few more supported mice
HID: logitech-dj: rename "gaming" receiver to "lightspeed"
Denis Kirjanov [Tue, 6 Aug 2019 10:51:11 +0000 (12:51 +0200)]
be2net: disable bh with spin_lock in be_process_mcc
be_process_mcc() is invoked in 3 different places and
always with BHs disabled except the be_poll function
but since it's invoked from softirq with BHs
disabled it won't hurt.
v1->v2: added explanation to the patch
v2->v3: add a missing call from be_cmds.c
Chen-Yu Tsai [Tue, 6 Aug 2019 07:35:39 +0000 (15:35 +0800)]
net: ethernet: sun4i-emac: Support phy-handle property for finding PHYs
The sun4i-emac uses the "phy" property to find the PHY it's supposed to
use. This property was deprecated in favor of "phy-handle" in commit 8c5b09447625 ("dt-bindings: net: sun4i-emac: Convert the binding to a
schemas").
Add support for this new property name, and fall back to the old one in
case the device tree hasn't been updated.
Linus Torvalds [Tue, 6 Aug 2019 18:22:22 +0000 (11:22 -0700)]
Merge branch 'x86/grand-schemozzle' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull pti updates from Thomas Gleixner:
"The performance deterioration departement is not proud at all to
present yet another set of speculation fences to mitigate the next
chapter in the 'what could possibly go wrong' story.
The new vulnerability belongs to the Spectre class and affects GS
based data accesses and has therefore been dubbed 'Grand Schemozzle'
for secret communication purposes. It's officially listed as
CVE-2019-1125.
Conditional branches in the entry paths which contain a SWAPGS
instruction (interrupts and exceptions) can be mis-speculated which
results in speculative accesses with a wrong GS base.
This can happen on entry from user mode through a mis-speculated
branch which takes the entry from kernel mode path and therefore does
not execute the SWAPGS instruction. The following speculative accesses
are done with user GS base.
On entry from kernel mode the mis-speculated branch executes the
SWAPGS instruction in the entry from user mode path which has the same
effect that the following GS based accesses are done with user GS
base.
If there is a disclosure gadget available in these code paths the
mis-speculated data access can be leaked through the usual side
channels.
The entry from user mode issue affects all CPUs which have speculative
execution. The entry from kernel mode issue affects only Intel CPUs
which can speculate through SWAPGS. On CPUs from other vendors SWAPGS
has semantics which prevent that.
SMAP migitates both problems but only when the CPU is not affected by
the Meltdown vulnerability.
The mitigation is to issue LFENCE instructions in the entry from
kernel mode path for all affected CPUs and on the affected Intel CPUs
also in the entry from user mode path unless PTI is enabled because
the CR3 write is serializing.
The fences are as usual enabled conditionally and can be completely
disabled on the kernel command line. The Spectre V1 documentation is
updated accordingly.
A big "Thank You!" goes to Josh for doing the heavy lifting for this
round of hardware misfeature 'repair'. Of course also "Thank You!" to
everybody else who contributed in one way or the other"
* 'x86/grand-schemozzle' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
Documentation: Add swapgs description to the Spectre v1 documentation
x86/speculation/swapgs: Exclude ATOMs from speculation through SWAPGS
x86/entry/64: Use JMP instead of JMPQ
x86/speculation: Enable Spectre v1 swapgs mitigations
x86/speculation: Prepare entry code for Spectre v1 swapgs mitigations
Kevin Hao [Fri, 26 Jul 2019 02:30:48 +0000 (10:30 +0800)]
mmc: cavium: Set the correct dma max segment size for mmc_host
We have set the mmc_host.max_seg_size to 8M, but the dma max segment
size of PCI device is set to 64K by default in function pci_device_add().
The mmc_host.max_seg_size is used to set the max segment size of
the blk queue. Then this mismatch will trigger a calltrace like below
when a bigger than 64K segment request arrives at mmc dev. So we should
consider the limitation of the cvm_mmc_host when setting the
mmc_host.max_seg_size.
DMA-API: thunderx_mmc 0000:01:01.4: mapping sg segment longer than device claims to support [len=131072] [max=65536]
WARNING: CPU: 6 PID: 238 at kernel/dma/debug.c:1221 debug_dma_map_sg+0x2b8/0x350
Modules linked in:
CPU: 6 PID: 238 Comm: kworker/6:1H Not tainted 5.3.0-rc1-next-20190724-yocto-standard+ #62
Hardware name: Marvell OcteonTX CN96XX board (DT)
Workqueue: kblockd blk_mq_run_work_fn
pstate: 80c00009 (Nzcv daif +PAN +UAO)
pc : debug_dma_map_sg+0x2b8/0x350
lr : debug_dma_map_sg+0x2b8/0x350
sp : ffff00001770f9e0
x29: ffff00001770f9e0 x28: ffffffff00000000
x27: 00000000ffffffff x26: ffff800bc2c73180
x25: ffff000010e83700 x24: 0000000000000002
x23: 0000000000000001 x22: 0000000000000001
x21: 0000000000000000 x20: ffff800bc48ba0b0
x19: ffff800bc97e8c00 x18: ffffffffffffffff
x17: 0000000000000000 x16: 0000000000000000
x15: ffff000010e835c8 x14: 6874207265676e6f
x13: 6c20746e656d6765 x12: 7320677320676e69
x11: 7070616d203a342e x10: 31303a31303a3030
x9 : 303020636d6d5f78 x8 : 35363d78616d5b20
x7 : 00000000000002fd x6 : ffff000010fd57dc
x5 : 0000000000000000 x4 : ffff0000106c61f0
x3 : 00000000ffffffff x2 : 0000800bee060000
x1 : 7010678df3041a00 x0 : 0000000000000000
Call trace:
debug_dma_map_sg+0x2b8/0x350
cvm_mmc_request+0x3c4/0x988
__mmc_start_request+0x9c/0x1f8
mmc_start_request+0x7c/0xb0
mmc_blk_mq_issue_rq+0x5c4/0x7b8
mmc_mq_queue_rq+0x11c/0x278
blk_mq_dispatch_rq_list+0xb0/0x568
blk_mq_do_dispatch_sched+0x6c/0x108
blk_mq_sched_dispatch_requests+0x110/0x1b8
__blk_mq_run_hw_queue+0xb0/0x118
blk_mq_run_work_fn+0x28/0x38
process_one_work+0x210/0x490
worker_thread+0x48/0x458
kthread+0x130/0x138
ret_from_fork+0x10/0x1c
Baolin Wang [Thu, 25 Jul 2019 03:14:22 +0000 (11:14 +0800)]
mmc: sdhci-sprd: Fix the incorrect soft reset operation when runtime resuming
The SD host controller specification defines 3 types software reset:
software reset for data line, software reset for command line and software
reset for all. Software reset for all means this reset affects the entire
Host controller except for the card detection circuit.
In sdhci_runtime_resume_host() we always do a software "reset for all",
which causes the Spreadtrum variant controller to work abnormally after
resuming. To fix the problem, let's do a software reset for the data and
the command part, rather than "for all".
However, as sdhci_runtime_resume() is a common sdhci function and we don't
want to change the behaviour for other variants, let's introduce a new
in-parameter for it. This enables the caller to decide if a "reset for all"
shall be done or not.
One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:
Lu Baolu [Thu, 1 Aug 2019 03:14:58 +0000 (11:14 +0800)]
iommu/vt-d: Detach domain when move device out of group
When removing a device from an iommu group, the domain should
be detached from the device. Otherwise, the stale domain info
will still be cached by the driver and the driver will refuse
to attach any domain to the device again.
Anders Roxell [Tue, 30 Jul 2019 15:20:11 +0000 (17:20 +0200)]
iommu/arm-smmu: Mark expected switch fall-through
Now that -Wimplicit-fallthrough is passed to GCC by default, the
following warning shows up:
../drivers/iommu/arm-smmu-v3.c: In function ‘arm_smmu_write_strtab_ent’:
../drivers/iommu/arm-smmu-v3.c:1189:7: warning: this statement may fall
through [-Wimplicit-fallthrough=]
if (disable_bypass)
^
../drivers/iommu/arm-smmu-v3.c:1191:3: note: here
default:
^~~~~~~
Rework so that the compiler doesn't warn about fall-through. Make it
clearer by calling 'BUG_ON()' when disable_bypass is set, and always
'break;'
Robin Murphy [Mon, 29 Jul 2019 15:32:38 +0000 (16:32 +0100)]
iommu/dma: Handle MSI mappings separately
MSI pages must always be mapped into a device's *current* domain, which
*might* be the default DMA domain, but might instead be a VFIO domain
with its own MSI cookie. This subtlety got accidentally lost in the
streamlining of __iommu_dma_map(), but rather than reintroduce more
complexity and/or special-casing, it turns out neater to just split this
path out entirely.
Since iommu_dma_get_msi_page() already duplicates much of what
__iommu_dma_map() does, it can easily just make the allocation and
mapping calls directly as well. That way we can further streamline the
helper back to exclusively operating on DMA domains.
Takashi Iwai [Tue, 6 Aug 2019 12:03:56 +0000 (14:03 +0200)]
ALSA: hda - Don't override global PCM hw info flag
The commit bfcba288b97f ("ALSA - hda: Add support for link audio time
reporting") introduced the conditional PCM hw info setup, but it
overwrites the global azx_pcm_hw object. This will cause a problem if
any other HD-audio controller, as it'll inherit the same bit flag
although another controller doesn't support that feature.
Fix the bug by setting the PCM hw info flag locally.
Fixes: bfcba288b97f ("ALSA - hda: Add support for link audio time reporting") Cc: <[email protected]> Signed-off-by: Takashi Iwai <[email protected]>
Vasily Gorbik [Mon, 5 Aug 2019 12:25:16 +0000 (14:25 +0200)]
s390: put _stext and _etext into .text section
Perf relies on _etext and _stext symbols being one of 't', 'T', 'v' or
'V'. Put them into .text section to guarantee that.
Also moves padding to page boundary inside .text which has an effect that
.text section is now padded with nops rather than 0's, which apparently
has been the initial intention for specifying 0x0700 fill expression.
Vasily Gorbik [Mon, 5 Aug 2019 13:07:19 +0000 (15:07 +0200)]
s390/unwind: remove stack recursion warning
Remove pointless stack recursion on stack type ... warning, which
only confuses people. There is no way to make backchain unwinder 100%
reliable. When a task is interrupted in-between stack frame allocation
and backchain write instructions new stack frame backchain pointer is
left uninitialized (there are also sometimes additional instruction
in-between stack frame allocation and backchain write instructions due
to gcc shrink-wrapping). In attempt to unwind such stack the unwinder
would still try to use that invalid backchain value and perform all kind
of sanity checks on it to make sure we are not pointed out of stack. In
some cases that invalid backchain value would be 0 and we would falsely
treat next stackframe as pt_regs and again gprs[15] in those pt_regs
might happen to point at some address within the task's stack.
Vasily Gorbik [Fri, 2 Aug 2019 15:44:29 +0000 (17:44 +0200)]
s390/setup: adjust start_code of init_mm to _text
After some investigation it doesn't look like init_mm fields
start_code/end_code are used anywhere besides potentially in dump_mm for
debugging purposes. Originally the value of 0 for start_code reflected
the presence of lowcore and early boot code. But with kaslr in place
start_code/end_code range should not span over unoccupied by the code
segment memory. So, adjust init_mm start_code to point at the beginning
of the code segment like other architectures do it.
s390/mm: fix dump_pagetables top level page table walking
Since commit d1874a0c2805 ("s390/mm: make the pxd_offset functions more
robust") behaviour of p4d_offset, pud_offset and pmd_offset has been
changed so that they cannot be used to iterate through top level page
table, because the index for the top level page table is now calculated
in pgd_offset. To avoid dumping the very first region/segment top level
table entry 2048 times simply iterate entry pointer like it is already
done in other page walking cases.
Fixes: d1874a0c2805 ("s390/mm: make the pxd_offset functions more robust") Reported-by: Ilya Leoshkevich <[email protected]> Reviewed-by: Heiko Carstens <[email protected]> Signed-off-by: Vasily Gorbik <[email protected]>
Vasily Gorbik [Fri, 2 Aug 2019 11:27:22 +0000 (13:27 +0200)]
s390/protvirt: avoid memory sharing for diag 308 set/store
This reverts commit db9492cef45e ("s390/protvirt: add memory sharing for
diag 308 set/store") which due to ultravisor implementation change is
not needed after all.
Fixes: db9492cef45e ("s390/protvirt: add memory sharing for diag 308 set/store") Reviewed-by: Janosch Frank <[email protected]> Signed-off-by: Vasily Gorbik <[email protected]>
sched/psi: Do not require setsched permission from the trigger creator
When a process creates a new trigger by writing into /proc/pressure/*
files, permissions to write such a file should be used to determine whether
the process is allowed to do so or not. Current implementation would also
require such a process to have setsched capability. Setting of psi trigger
thread's scheduling policy is an implementation detail and should not be
exposed to the user level. Remove the permission check by using _nocheck
version of the function.
Peter Zijlstra [Thu, 1 Aug 2019 10:41:31 +0000 (12:41 +0200)]
sched/psi: Reduce psimon FIFO priority
PSI defaults to a FIFO-99 thread, reduce this to FIFO-1.
FIFO-99 is the very highest priority available to SCHED_FIFO and
it not a suitable default; it would indicate the psi work is the
most important work on the machine.
Since Real-Time tasks will have pre-allocated memory and locked it in
place, Real-Time tasks do not care about PSI. All it needs is to be
above OTHER.
sched/deadline: Fix double accounting of rq/running bw in push & pull
{push,pull}_dl_task() always calls {de,}activate_task() with .flags=0
which sets p->on_rq=TASK_ON_RQ_MIGRATING.
{push,pull}_dl_task()->{de,}activate_task()->{de,en}queue_task()->
{de,en}queue_task_dl() calls {sub,add}_{running,rq}_bw() since
p->on_rq==TASK_ON_RQ_MIGRATING.
So {sub,add}_{running,rq}_bw() in {push,pull}_dl_task() is
double-accounting for that task.
Fix it by removing rq/running bw accounting in [push/pull]_dl_task().
HID: sony: Fix race condition between rumble and device remove.
Valve reported a kernel crash on Ubuntu 18.04 when disconnecting a DS4
gamepad while rumble is enabled. This issue is reproducible with a
frequency of 1 in 3 times in the game Borderlands 2 when using an
automatic weapon, which triggers many rumble operations.
We found the issue to be a race condition between sony_remove and the
final device destruction by the HID / input system. The problem was
that sony_remove didn't clean some of its work_item state in
"struct sony_sc". After sony_remove work, the corresponding evdev
node was around for sufficient time for applications to still queue
rumble work after "sony_remove".
On pre-4.19 kernels the race condition caused a kernel crash due to a
NULL-pointer dereference as "sc->output_report_dmabuf" got freed during
sony_remove. On newer kernels this crash doesn't happen due the buffer
now being allocated using devm_kzalloc. However we can still queue work,
while the driver is an undefined state.
This patch fixes the described problem, by guarding the work_item
"state_worker" with an initialized variable, which we are setting back
to 0 on cleanup.
Wenwen Wang [Tue, 6 Aug 2019 07:00:27 +0000 (03:00 -0400)]
ALSA: usb-audio: fix a memory leak bug
In snd_usb_get_audioformat_uac3(), a structure for channel maps 'chmap' is
allocated through kzalloc() before the execution goto 'found_clock'.
However, this structure is not deallocated if the memory allocation for
'pd' fails, leading to a memory leak bug.
To fix the above issue, free 'fp->chmap' before returning NULL.
==================================================================
BUG: KASAN: use-after-free in __lock_acquire+0x302a/0x3b50
kernel/locking/lockdep.c:3753
Read of size 8 at addr ffff8881cf591a08 by task syz-executor.1/26260
Memory state around the buggy address: ffff8881cf591900: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8881cf591980: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ffff8881cf591a00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^ ffff8881cf591a80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8881cf591b00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
==================================================================
In order to avoid opening a disconnected device, we need to check exist
again after acquiring the existance lock, and bail out if necessary.
projects / linux.git / log