xen/virtio: Convert PAGE_SIZE/PAGE_SHIFT/PFN_UP to Xen counterparts
Currently, a grant ref is always based on the Xen page granularity
(4KB), and guest commonly uses the same page granularity.
But the guest may use a different page granularity (i.e 64KB).
So adopt the code to be able to deal with it.
xen/virtio: Handle cases when page offset > PAGE_SIZE properly
Passed to xen_grant_dma_map_page() offset in the page
can be > PAGE_SIZE even if the guest uses the same page granularity
as Xen (4KB).
Before current patch, if such case happened we ended up providing
grants for the whole region in xen_grant_dma_map_page() which
was really unnecessary. The more, we ended up not releasing all
grants which represented that region in xen_grant_dma_unmap_page().
Current patch updates the code to be able to deal with such cases.
xen/pv: support selecting safe/unsafe msr accesses
Instead of always doing the safe variants for reading and writing MSRs
in Xen PV guests, make the behavior controllable via Kconfig option
and a boot parameter.
The default will be the current behavior, which is to always use the
safe variant.
xen/pv: refactor msr access functions to support safe and unsafe accesses
Refactor and rename xen_read_msr_safe() and xen_write_msr_safe() to
support both cases of MSR accesses, safe ones and potentially GP-fault
generating ones.
This will prepare to no longer swallow GPs silently in xen_read_msr()
and xen_write_msr().
Juergen Gross [Wed, 5 Oct 2022 07:42:33 +0000 (09:42 +0200)]
xen/pv: fix vendor checks for pmu emulation
The CPU vendor checks for pmu emulation are rather limited today, as
the assumption seems to be that only Intel and AMD are existing and/or
supported vendors.
Fix that by handling Centaur and Zhaoxin CPUs the same way as Intel,
and Hygon the same way as AMD.
While at it fix the return type of is_intel_pmu_msr().
Suggested-by: Jan Beulich <jbeulich@suse.com> Signed-off-by: Juergen Gross <jgross@suse.com> Reviewed-by: Jan Beulich <jbeulich@suse.com> Signed-off-by: Juergen Gross <jgross@suse.com>
xen/pv: add fault recovery control to pmu msr accesses
Today pmu_msr_read() and pmu_msr_write() fall back to the safe variants
of read/write MSR in case the MSR access isn't emulated via Xen. Allow
the caller to select that faults should not be recovered from by passing
NULL for the error pointer.
Juergen Gross [Mon, 29 Aug 2022 11:26:07 +0000 (13:26 +0200)]
xen/virtio: restructure xen grant dma setup
In order to prepare supporting other means than device tree for
setting up virtio devices under Xen, restructure the functions
xen_is_grant_dma_device() and xen_grant_setup_dma_ops() a little bit.
Juergen Gross [Thu, 6 Oct 2022 08:50:28 +0000 (10:50 +0200)]
xen/pcifront: move xenstore config scanning into sub-function
pcifront_try_connect() and pcifront_attach_devices() share a large
chunk of duplicated code for reading the config information from
Xenstore, which only differs regarding calling pcifront_rescan_root()
or pcifront_scan_root().
Put that code into a new sub-function. It is fine to always call
pcifront_rescan_root() from that common function, as it will fallback
to pcifront_scan_root() if the domain/bus combination isn't known
yet (and pcifront_scan_root() should never be called for an already
known domain/bus combination anyway). In order to avoid duplicate
messages for the fallback case move the check for domain/bus not known
to the beginning of pcifront_rescan_root().
While at it fix the error reporting in case the root-xx node had the
wrong format.
As the return value of pcifront_try_connect() and
pcifront_attach_devices() are not used anywhere make those functions
return void. As an additional bonus this removes the dubious return
of -EFAULT in case of an unexpected driver state.
M. Vefa Bicakci [Sun, 2 Oct 2022 22:20:06 +0000 (18:20 -0400)]
xen/gntdev: Accommodate VMA splitting
Prior to this commit, the gntdev driver code did not handle the
following scenario correctly with paravirtualized (PV) Xen domains:
* User process sets up a gntdev mapping composed of two grant mappings
(i.e., two pages shared by another Xen domain).
* User process munmap()s one of the pages.
* User process munmap()s the remaining page.
* User process exits.
In the scenario above, the user process would cause the kernel to log
the following messages in dmesg for the first munmap(), and the second
munmap() call would result in similar log messages:
For each munmap() call, the Xen hypervisor (if built with CONFIG_DEBUG)
would print out the following and trigger a general protection fault in
the affected Xen PV domain:
(XEN) d0v... Attempt to implicitly unmap d0's grant PTE ...
(XEN) d0v... Attempt to implicitly unmap d0's grant PTE ...
As of this writing, gntdev_grant_map structure's vma field (referred to
as map->vma below) is mainly used for checking the start and end
addresses of mappings. However, with split VMAs, these may change, and
there could be more than one VMA associated with a gntdev mapping.
Hence, remove the use of map->vma and rely on map->pages_vm_start for
the original start address and on (map->count << PAGE_SHIFT) for the
original mapping size. Let the invalidate() and find_special_page()
hooks use these.
Also, given that there can be multiple VMAs associated with a gntdev
mapping, move the "mmu_interval_notifier_remove(&map->notifier)" call to
the end of gntdev_put_map, so that the MMU notifier is only removed
after the closing of the last remaining VMA.
Finally, use an atomic to prevent inadvertent gntdev mapping re-use,
instead of using the map->live_grants atomic counter and/or the map->vma
pointer (the latter of which is now removed). This prevents the
userspace from mmap()'ing (with MAP_FIXED) a gntdev mapping over the
same address range as a previously set up gntdev mapping. This scenario
can be summarized with the following call-trace, which was valid prior
to this commit:
mmap
gntdev_mmap
mmap (repeat mmap with MAP_FIXED over the same address range)
gntdev_invalidate
unmap_grant_pages (sets 'being_removed' entries to true)
gnttab_unmap_refs_async
unmap_single_vma
gntdev_mmap (maps the shared pages again)
munmap
gntdev_invalidate
unmap_grant_pages
(no-op because 'being_removed' entries are true)
unmap_single_vma (For PV domains, Xen reports that a granted page
is being unmapped and triggers a general protection fault in the
affected domain, if Xen was built with CONFIG_DEBUG)
The fix for this last scenario could be worth its own commit, but we
opted for a single commit, because removing the gntdev_grant_map
structure's vma field requires guarding the entry to gntdev_mmap(), and
the live_grants atomic counter is not sufficient on its own to prevent
the mmap() over a pre-existing mapping.
M. Vefa Bicakci [Sun, 2 Oct 2022 22:20:05 +0000 (18:20 -0400)]
xen/gntdev: Prevent leaking grants
Prior to this commit, if a grant mapping operation failed partially,
some of the entries in the map_ops array would be invalid, whereas all
of the entries in the kmap_ops array would be valid. This in turn would
cause the following logic in gntdev_map_grant_pages to become invalid:
for (i = 0; i < map->count; i++) {
if (map->map_ops[i].status == GNTST_okay) {
map->unmap_ops[i].handle = map->map_ops[i].handle;
if (!use_ptemod)
alloced++;
}
if (use_ptemod) {
if (map->kmap_ops[i].status == GNTST_okay) {
if (map->map_ops[i].status == GNTST_okay)
alloced++;
map->kunmap_ops[i].handle = map->kmap_ops[i].handle;
}
}
}
...
atomic_add(alloced, &map->live_grants);
Assume that use_ptemod is true (i.e., the domain mapping the granted
pages is a paravirtualized domain). In the code excerpt above, note that
the "alloced" variable is only incremented when both kmap_ops[i].status
and map_ops[i].status are set to GNTST_okay (i.e., both mapping
operations are successful). However, as also noted above, there are
cases where a grant mapping operation fails partially, breaking the
assumption of the code excerpt above.
The aforementioned causes map->live_grants to be incorrectly set. In
some cases, all of the map_ops mappings fail, but all of the kmap_ops
mappings succeed, meaning that live_grants may remain zero. This in turn
makes it impossible to unmap the successfully grant-mapped pages pointed
to by kmap_ops, because unmap_grant_pages has the following snippet of
code at its beginning:
if (atomic_read(&map->live_grants) == 0)
return; /* Nothing to do */
In other cases where only some of the map_ops mappings fail but all
kmap_ops mappings succeed, live_grants is made positive, but when the
user requests unmapping the grant-mapped pages, __unmap_grant_pages_done
will then make map->live_grants negative, because the latter function
does not check if all of the pages that were requested to be unmapped
were actually unmapped, and the same function unconditionally subtracts
"data->count" (i.e., a value that can be greater than map->live_grants)
from map->live_grants. The side effects of a negative live_grants value
have not been studied.
The net effect of all of this is that grant references are leaked in one
of the above conditions. In Qubes OS v4.1 (which uses Xen's grant
mechanism extensively for X11 GUI isolation), this issue manifests
itself with warning messages like the following to be printed out by the
Linux kernel in the VM that had granted pages (that contain X11 GUI
window data) to dom0: "g.e. 0x1234 still pending", especially after the
user rapidly resizes GUI VM windows (causing some grant-mapping
operations to partially or completely fail, due to the fact that the VM
unshares some of the pages as part of the window resizing, making the
pages impossible to grant-map from dom0).
The fix for this issue involves counting all successful map_ops and
kmap_ops mappings separately, and then adding the sum to live_grants.
During unmapping, only the number of successfully unmapped grants is
subtracted from live_grants. The code is also modified to check for
negative live_grants values after the subtraction and warn the user.
xen/virtio: Fix potential deadlock when accessing xen_grant_dma_devices
As find_xen_grant_dma_data() is called from both interrupt and process
contexts, the access to xen_grant_dma_devices XArray must be protected
by xa_lock_irqsave to avoid deadlock scenario.
As XArray API doesn't provide xa_store_irqsave helper, call lockless
__xa_store directly and guard it externally.
Also move the storage of the XArray's entry to a separate helper.
Jason Andryuk [Mon, 29 Aug 2022 15:15:36 +0000 (11:15 -0400)]
xen-pcifront: Handle missed Connected state
An HVM guest with linux stubdomain and 2 PCI devices failed to start as
libxl timed out waiting for the PCI devices to be added. It happens
intermittently but with some regularity. libxl wrote the two xenstore
entries for the devices, but then timed out waiting for backend state 4
(Connected) - the state stayed at 7 (Reconfiguring). (PCI passthrough
to an HVM with stubdomain is PV passthrough to the stubdomain and then
HVM passthrough with the QEMU inside the stubdomain.)
The stubdomain kernel never printed "pcifront pci-0: Installing PCI
frontend", so it seems to have missed state 4 which would have
called pcifront_try_connect() -> pcifront_connect_and_init_dma()
Have pcifront_detach_devices() special-case state Initialised and call
pcifront_connect_and_init_dma(). Don't use pcifront_try_connect()
because that sets the xenbus state which may throw off the backend.
After connecting, skip the remainder of detach_devices since none have
been initialized yet. When the backend switches to Reconfigured,
pcifront_attach_devices() will pick them up again.
Linus Torvalds [Sun, 2 Oct 2022 16:41:27 +0000 (09:41 -0700)]
Merge tag 'perf-urgent-2022-10-02' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull misc perf fixes from Ingo Molnar:
- Fix a PMU enumeration/initialization bug on Intel Alder Lake CPUs
- Fix KVM guest PEBS register handling
- Fix race/reentry bug in perf_output_read_group() reading of PMU
counters
* tag 'perf-urgent-2022-10-02' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/core: Fix reentry problem in perf_output_read_group()
perf/x86/core: Completely disable guest PEBS via guest's global_ctrl
perf/x86/intel: Fix unchecked MSR access error for Alder Lake N
Linus Torvalds [Sun, 2 Oct 2022 16:30:35 +0000 (09:30 -0700)]
Merge tag 'x86_urgent_for_v6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull x86 fixes from Borislav Petkov:
- Add the respective UP last level cache mask accessors in order not to
cause segfaults when lscpu accesses their representation in sysfs
- Fix for a race in the alternatives batch patching machinery when
kprobes are set
* tag 'x86_urgent_for_v6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/cacheinfo: Add a cpu_llc_shared_mask() UP variant
x86/alternative: Fix race in try_get_desc()
i2c: davinci: fix PM disable depth imbalance in davinci_i2c_probe
The pm_runtime_enable will increase power disable depth. Thus a
pairing decrement is needed on the error handling path to keep
it balanced according to context.
Linus Torvalds [Sat, 1 Oct 2022 16:39:42 +0000 (09:39 -0700)]
Merge tag 'usb-6.0-final' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Pull USB/Thunderbolt fixes from Greg KH:
"Here are some tiny USB and Thunderbolt driver fixes and quirks.
Included in here are:
- three uas/usb-storage driver quirks to get the devices working
properly due to broken firmware images in them (they can not run at
high data rates, and are also throttled on other operating systems
because of this)
- thunderbolt bugfix for plug event delays
- typec runtime warning removal
- dwc3 st driver bugfix. Note, a follow-on fix for this will end up
coming in for 6.1-rc1 as the developers are still arguing over what
the final solution will be, but this should be sufficient for now
All of these have been in linux-next with no reported problems"
* tag 'usb-6.0-final' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
uas: ignore UAS for Thinkplus chips
usb-storage: Add Hiksemi USB3-FW to IGNORE_UAS
uas: add no-uas quirk for Hiksemi usb_disk
usb: dwc3: st: Fix node's child name
usb: typec: ucsi: Remove incorrect warning
thunderbolt: Explicitly reset plug events delay back to USB4 spec value
Linus Torvalds [Sat, 1 Oct 2022 16:27:18 +0000 (09:27 -0700)]
Merge tag 'media/v6.0-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media
Pull media fixes from Mauro Carvalho Chehab:
- some fixes for the v4l2 ioctl handler logic
- a fix for an out of bound access in the DVB videobuf2 handler
- three driver fixes (rkvdec, mediatek/vcodek and uvcvideo)
* tag 'media/v6.0-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:
media: rkvdec: Disable H.264 error detection
media: mediatek: vcodec: Drop platform_get_resource(IORESOURCE_IRQ)
media: dvb_vb2: fix possible out of bound access
media: v4l2-ioctl.c: fix incorrect error path
media: v4l2-compat-ioctl32.c: zero buffer passed to v4l2_compat_get_array_args()
media: uvcvideo: Fix InterfaceProtocol for Quanta camera
Linus Torvalds [Sat, 1 Oct 2022 16:13:29 +0000 (09:13 -0700)]
Merge tag 'mm-hotfixes-stable-2022-09-30' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull more hotfixes from Andrew Morton:
"One MAINTAINERS update, two MM fixes, both cc:stable"
The previous pull wasn't fated to be the last one..
* tag 'mm-hotfixes-stable-2022-09-30' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm:
damon/sysfs: fix possible memleak on damon_sysfs_add_target
mm: fix BUG splat with kvmalloc + GFP_ATOMIC
MAINTAINERS: drop entry to removed file in ARM/RISCPC ARCHITECTURE
Levi Yun [Mon, 26 Sep 2022 16:06:11 +0000 (16:06 +0000)]
damon/sysfs: fix possible memleak on damon_sysfs_add_target
When damon_sysfs_add_target couldn't find proper task, New allocated
damon_target structure isn't registered yet, So, it's impossible to free
new allocated one by damon_sysfs_destroy_targets.
By calling damon_add_target as soon as allocating new target, Fix this
possible memory leak.
Link: https://lkml.kernel.org/r/20220926160611.48536-1-sj@kernel.org Fixes: a61ea561c871 ("mm/damon/sysfs: link DAMON for virtual address spaces monitoring") Signed-off-by: Levi Yun <ppbuk5246@gmail.com> Signed-off-by: SeongJae Park <sj@kernel.org> Reviewed-by: SeongJae Park <sj@kernel.org> Cc: <stable@vger.kernel.org> [5.17.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
MAINTAINERS: drop entry to removed file in ARM/RISCPC ARCHITECTURE
Commit c1fe8d054c0a ("ARM: riscpc: use GENERIC_IRQ_MULTI_HANDLER") removes
arch/arm/include/asm/hardware/entry-macro-iomd.S, but missed to adjust
MAINTAINERS.
Hence, ./scripts/get_maintainer.pl --self-test=patterns complains about a
broken reference.
Drop the file entry to the removed file in ARM/RISCPC ARCHITECTURE.
Link: https://lkml.kernel.org/r/20220919075255.386-1-lukas.bulwahn@gmail.com Fixes: c1fe8d054c0a ("ARM: riscpc: use GENERIC_IRQ_MULTI_HANDLER") Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com> Cc: Ard Biesheuvel <ardb@kernel.org> Cc: Arnd Bergmann <arnd@arndb.de> Cc: Russell King <linux@armlinux.org.uk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Merge tag 'drm-fixes-2022-10-01' of git://anongit.freedesktop.org/drm/drm
Pull drm fixes from Daniel Vetter:
"Some last minute amd fixes:
- VCN 4.x and GC 11.x fixes, mostly around fw"
* tag 'drm-fixes-2022-10-01' of git://anongit.freedesktop.org/drm/drm:
drm/amdgpu/gfx11: switch to amdgpu_gfx_rlc_init_microcode
drm/amdgpu: add helper to init rlc firmware
drm/amdgpu: add helper to init rlc fw in header v2_4
drm/amdgpu: add helper to init rlc fw in header v2_3
drm/amdgpu: add helper to init rlc fw in header v2_2
drm/amdgpu: add helper to init rlc fw in header v2_1
drm/amdgpu: add helper to init rlc fw in header v2_0
drm/amdgpu: save rlcv/rlcp ucode version in amdgpu_gfx
drm/amdgpu: Enable sram on vcn_4_0_2
drm/amdgpu: Enable VCN DPG for GC11_0_1
Merge tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk driver fixes from Stephen Boyd:
"Here's the last batch of clk driver fixes for this release.
These patches fix serious problems, for example, i.MX has an issue
where changing the NAND clk frequency hangs the system. On Allwinner
H6 the GPU is being overclocked which could lead to long term hardware
damage.
And finally on some Broadcom SoCs the serial console stopped working
because the clk tree hierarchy description got broken by an
inadvertant DT node name change. That's fixed by using
'clock-output-names' to generate a stable and unique name for clks so
the framework can properly link things up.
There's also a couple build fixes in here. One to fix CONFIG_OF=n
builds and one to avoid an array out of bounds bug that happens during
clk registration on microchip. I hope that KASAN would have found that
OOB problem, but probably KASAN wasn't attempted. Instead LLVM/clang
compilation caused an oops, while GCC didn't"
* tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
clk: imx93: drop of_match_ptr
clk: iproc: Do not rely on node name for correct PLL setup
clk: sunxi-ng: h6: Fix default PLL GPU rate
clk: imx: imx6sx: remove the SET_RATE_PARENT flag for QSPI clocks
clk: microchip: mpfs: make the rtc's ahb clock critical
clk: microchip: mpfs: fix clk_cfg array bounds violation
clk: ingenic-tcu: Properly enable registers before accessing timers
Merge tag 'perf-tools-fixes-for-v6.0-2022-09-29' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux
Pull perf tools fixes from Arnaldo Carvalho de Melo:
- Fail the 'perf test record' entry on error, fixing a regression where
just setup stuff like allocating memory and not the actual things
being tested failed.
- Fixup disabling of -Wdeprecated-declarations for the python scripting
engine, the previous attempt had a brown paper bag thinko.
- Fix branch stack sampling test to include sanity check for branch
filter on PowerPC.
- Update is_ignored_symbol function to match the kernel ignored list,
fixing running the 'perf test' entry that compares resolving symbols
from kallsyms to resolving from vmlinux.
- Augment the data source type with ARM's neoverse_spe list, the
previous code was limited in its search resolving the data source.
- Fix some clang 5 variable set but unused cases.
- Get a perf cgroup more portably in BPF as the
__builtin_preserve_enum_value builtin is not available in older
versions of clang. In those cases we can forgo BPF's CO-RE (Compile
Once, Run Everywhere).
- More Fixes for Intel's hybrid CPU model.
* tag 'perf-tools-fixes-for-v6.0-2022-09-29' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
perf build: Fixup disabling of -Wdeprecated-declarations for the python scripting engine
perf tests mmap-basic: Remove unused variable to address clang 15 warning
perf parse-events: Ignore clang 15 warning about variable set but unused in bison produced code
perf tests record: Fail the test if the 'errs' counter is not zero
perf test: Fix test case 87 ("perf record tests") for hybrid systems
perf arm-spe: augment the data source type with neoverse_spe list
perf tests vmlinux-kallsyms: Update is_ignored_symbol function to match the kernel ignored list
perf tests powerpc: Fix branch stack sampling test to include sanity check for branch filter
perf parse-events: Remove "not supported" hybrid cache events
perf print-events: Fix "perf list" can not display the PMU prefix for some hybrid cache events
perf tools: Get a perf cgroup more portably in BPF
Merge tag 'for-linus-6.0' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull kvm fixes from Paolo Bonzini:
"A small fix to the reported set of supported CPUID bits, and selftests
fixes:
- Skip tests that require EPT when it is not available
- Do not hang when a test fails with an empty stack trace
- avoid spurious failure when running access_tracking_perf_test in a
KVM guest
- work around GCC's tendency to optimize loops into mem*() functions,
which breaks because the guest code in selftests cannot call into
PLTs
- fix -Warray-bounds error in fix_hypercall_test"
* tag 'for-linus-6.0' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: selftests: Compare insn opcodes directly in fix_hypercall_test
KVM: selftests: Implement memcmp(), memcpy(), and memset() for guest use
KVM: x86: Hide IA32_PLATFORM_DCA_CAP[31:0] from the guest
KVM: selftests: Gracefully handle empty stack traces
KVM: selftests: replace assertion with warning in access_tracking_perf_test
KVM: selftests: Skip tests that require EPT when it is not available
drm/amdgpu/gfx11: switch to amdgpu_gfx_rlc_init_microcode
switch to common helper to initialize rlc firmware
for gfx11
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
To initialzie rlc firmware according to rlc
firmware header version
v2: squash in backwards compat fix
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add helper to init rlc fw in header v2_4
To initialize rlc firmware in header v2_4
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add helper to init rlc fw in header v2_3
To initialize rlc firmware in header v2_3
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add helper to init rlc fw in header v2_2
To initialize rlc firmware in header v2_2
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add helper to init rlc fw in header v2_1
To initialize rlc firmware in header v2_1
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: add helper to init rlc fw in header v2_0
To initialize rlc firmware in header v2_0
Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Merge tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost
Pull virtio fixes from Michael Tsirkin:
"Some last minute fixes.
The virtio-blk one is the most important one since it was actually
seen in the field, but the rest of them are small and clearly safe,
everything here has been in next for a while"
* tag 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:
vdpa/mlx5: Fix MQ to support non power of two num queues
vduse: prevent uninitialized memory accesses
virtio-blk: Fix WARN_ON_ONCE in virtio_queue_rq()
virtio_test: fixup for vq reset
virtio-crypto: fix memory-leak
vdpa/ifcvf: fix the calculation of queuepair
Merge tag 'block-6.0-2022-09-29' of git://git.kernel.dk/linux
Pull block fixes from Jens Axboe:
"A single NVMe pull request via Christoph with a few fixes that should
go into the 6.0 release:
- Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices
(Michael Kelley)
- Disable Write Zeroes on Phison E3C/E4C (Tina Hsu)"
* tag 'block-6.0-2022-09-29' of git://git.kernel.dk/linux:
nvme-pci: disable Write Zeroes on Phison E3C/E4C
nvme: Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices
Merge tag 'io_uring-6.0-2022-09-29' of git://git.kernel.dk/linux
Pull io_uring fixes from Jens Axboe:
"Two fixes that should go into 6.0:
- Tweak the single issuer logic to register the task at creation,
rather than at first submit. SINGLE_ISSUER was added for 6.0, and
after some discussion on this, we decided to make it a bit stricter
while it's still possible to do so (Dylan).
- Stefan from Samba had some doubts on the level triggered poll that
was added for this release. Rather than attempt to mess around with
it now, just do the quick one-liner to disable it for release and
we have time to discuss and change it for 6.1 instead (me)"
* tag 'io_uring-6.0-2022-09-29' of git://git.kernel.dk/linux:
io_uring/poll: disable level triggered poll
io_uring: register single issuer task at creation
Merge tag 'pstore-v6.0-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux
Pull pstore revert from Kees Cook:
"A misbehavior with some compression backends in pstore was just
discovered due to the recent crypto acomp migration.
Since we're so close to release, it seems better to just simply revert
it, and we can figure out what's going on without leaving it broken
for a release.
- Revert crypto acomp migration (Guilherme G. Piccoli)"
* tag 'pstore-v6.0-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
Revert "pstore: migrate to crypto acomp interface"
When using this new interface, both efi_pstore and ramoops
backends are unable to properly decompress dmesg if using
zstd, lz4 and lzo algorithms (and maybe more). It does succeed
with deflate though.
The message observed in the kernel log is:
[2.328828] pstore: crypto_acomp_decompress failed, ret = -22!
The pstore infrastructure is able to collect the dmesg with
both backends tested, but since decompression fails it's
unreadable. With this revert everything is back to normal.
Merge tag 'drm-fixes-2022-09-30-1' of git://anongit.freedesktop.org/drm/drm
Pull drm fixes from Dave Airlie:
"Last set of fixes for 6.0 hopefully - minor bridge fixes, i915 fixes,
and a bunch of amdgpu fixes for new IP blocks, along with a couple of
regression fixes. Should be all set for merge window next week.
* tag 'drm-fixes-2022-09-30-1' of git://anongit.freedesktop.org/drm/drm: (26 commits)
drm/amd/display: Prevent OTG shutdown during PSR SU
drm/i915/gt: Perf_limit_reasons are only available for Gen11+
drm/amdgpu: Add amdgpu suspend-resume code path under SRIOV
drm/amdgpu: Remove fence_process in count_emitted
drm/amdgpu: Correct the position in patch_cond_exec
drm/amd/display: fill in clock values when DPM is not enabled
drm/amd/display: Avoid unnecessary pixel rate divider programming
drm/amd/display: Remove assert for odm transition case
drm/amd/display: Fix typo in get_pixel_rate_div
drm/amd/display: Fix audio on display after unplugging another
drm/amd/display: Add explicit FIFO disable for DP blank
drm/amd/display: Wrap OTG disable workaround with FIFO control
drm/amd/display: Do DIO FIFO enable after DP video stream enable
drm/amd/display: Update DCN32 to use new SR latencies
drm/amd/display: Avoid avoid unnecessary pixel rate divider programming
drm/amdkfd: fix dropped interrupt in kfd_int_process_v11
drm/amdgpu: pass queue size and is_aql_queue to MES
drm/amdkfd: fix MQD init for GFX11 in init_mqd
drm/amd/pm: use adverse selection for dpm features unsupported by driver
drm/amd/pm: enable gfxoff feature for SMU 13.0.0
...
KVM: selftests: Compare insn opcodes directly in fix_hypercall_test
Directly compare the expected versus observed hypercall instructions when
verifying that KVM patched in the native hypercall (FIX_HYPERCALL_INSN
quirk enabled). gcc rightly complains that doing a 4-byte memcpy() with
an "unsigned char" as the source generates an out-of-bounds accesses.
Alternatively, "exp" and "obs" could be declared as 3-byte arrays, but
there's no known reason to copy locally instead of comparing directly.
In function ‘assert_hypercall_insn’,
inlined from ‘guest_main’ at x86_64/fix_hypercall_test.c:91:2:
x86_64/fix_hypercall_test.c:63:9: error: array subscript ‘unsigned int[0]’
is partly outside array bounds of ‘unsigned char[1]’ [-Werror=array-bounds]
63 | memcpy(&exp, exp_insn, sizeof(exp));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
x86_64/fix_hypercall_test.c: In function ‘guest_main’:
x86_64/fix_hypercall_test.c:42:22: note: object ‘vmx_hypercall_insn’ of size 1
42 | extern unsigned char vmx_hypercall_insn;
| ^~~~~~~~~~~~~~~~~~
x86_64/fix_hypercall_test.c:25:22: note: object ‘svm_hypercall_insn’ of size 1
25 | extern unsigned char svm_hypercall_insn;
| ^~~~~~~~~~~~~~~~~~
In function ‘assert_hypercall_insn’,
inlined from ‘guest_main’ at x86_64/fix_hypercall_test.c:91:2:
x86_64/fix_hypercall_test.c:64:9: error: array subscript ‘unsigned int[0]’
is partly outside array bounds of ‘unsigned char[1]’ [-Werror=array-bounds]
64 | memcpy(&obs, obs_insn, sizeof(obs));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
x86_64/fix_hypercall_test.c: In function ‘guest_main’:
x86_64/fix_hypercall_test.c:25:22: note: object ‘svm_hypercall_insn’ of size 1
25 | extern unsigned char svm_hypercall_insn;
| ^~~~~~~~~~~~~~~~~~
x86_64/fix_hypercall_test.c:42:22: note: object ‘vmx_hypercall_insn’ of size 1
42 | extern unsigned char vmx_hypercall_insn;
| ^~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
make: *** [../lib.mk:135: tools/testing/selftests/kvm/x86_64/fix_hypercall_test] Error 1
Fixes: 6c2fa8b20d0c ("selftests: KVM: Test KVM_X86_QUIRK_FIX_HYPERCALL_INSN") Cc: Oliver Upton <oliver.upton@linux.dev> Signed-off-by: Sean Christopherson <seanjc@google.com> Reviewed-by: Oliver Upton <oliver.upton@linux.dev>
Message-Id: <20220928233652.783504-3-seanjc@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: selftests: Implement memcmp(), memcpy(), and memset() for guest use
Implement memcmp(), memcpy(), and memset() to override the compiler's
built-in versions in order to guarantee that the compiler won't generate
out-of-line calls to external functions via the PLT. This allows the
helpers to be safely used in guest code, as KVM selftests don't support
dynamic loading of guest code.
Steal the implementations from the kernel's generic versions, sans the
optimizations in memcmp() for unaligned accesses.
Put the utilities in a separate compilation unit and build with
-ffreestanding to fudge around a gcc "feature" where it will optimize
memset(), memcpy(), etc... by generating a recursive call. I.e. the
compiler optimizes itself into infinite recursion. Alternatively, the
individual functions could be tagged with
optimize("no-tree-loop-distribute-patterns"), but using "optimize" for
anything but debug is discouraged, and Linus NAK'd the use of the flag
in the kernel proper[*].
Cc: Andrew Jones <andrew.jones@linux.dev> Cc: Anup Patel <anup@brainfault.org> Cc: Atish Patra <atishp@atishpatra.org> Cc: Christian Borntraeger <borntraeger@linux.ibm.com> Cc: Janosch Frank <frankja@linux.ibm.com> Cc: Claudio Imbrenda <imbrenda@linux.ibm.com> Signed-off-by: Sean Christopherson <seanjc@google.com>
Message-Id: <20220928233652.783504-2-seanjc@google.com> Reviewed-by: Andrew Jones <andrew.jones@linux.dev> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Jim Mattson [Thu, 22 Sep 2022 23:18:54 +0000 (16:18 -0700)]
KVM: x86: Hide IA32_PLATFORM_DCA_CAP[31:0] from the guest
The only thing reported by CPUID.9 is the value of
IA32_PLATFORM_DCA_CAP[31:0] in EAX. This MSR doesn't even exist in the
guest, since CPUID.1:ECX.DCA[bit 18] is clear in the guest.
Clear CPUID.9 in KVM_GET_SUPPORTED_CPUID.
Fixes: 24c82e576b78 ("KVM: Sanitize cpuid") Signed-off-by: Jim Mattson <jmattson@google.com>
Message-Id: <20220922231854.249383-1-jmattson@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Bail out of test_dump_stack() if the stack trace is empty rather than
invoking addr2line with zero addresses. The problem with the latter is
that addr2line will block waiting for addresses to be passed in via
stdin, e.g. if running a selftest from an interactive terminal.
Opportunistically fix up the comment that mentions skipping 3 frames
since only 2 are skipped in the code.
Cc: Vipin Sharma <vipinsh@google.com> Cc: Sean Christopherson <seanjc@google.com> Signed-off-by: David Matlack <dmatlack@google.com>
Message-Id: <20220922231724.3560211-1-dmatlack@google.com>
[Small tweak to keep backtrace() call close to if(). - Paolo] Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
KVM: selftests: replace assertion with warning in access_tracking_perf_test
Page_idle uses {ptep/pmdp}_clear_young_notify which in turn calls
the mmu notifier callback ->clear_young(), which purposefully
does not flush the TLB.
When running the test in a nested guest, point 1. of the test
doc header is violated, because KVM TLB is unbounded by size
and since no flush is forced, KVM does not update the sptes
accessed/idle bits resulting in guest assertion failure.
More precisely, only the first ACCESS_WRITE in run_test() actually
makes visible changes, because sptes are created and the accessed
bit is set to 1 (or idle bit is 0). Then the first mark_memory_idle()
passes since access bit is still one, and sets all pages as idle
(or not accessed). When the next write is performed, the update
is not flushed therefore idle is still 1 and next mark_memory_idle()
fails.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20220926082923.299554-1-eesposit@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
perf build: Fixup disabling of -Wdeprecated-declarations for the python scripting engine
A brown paper bag where -Wno-error=deprecated-declarations was added
from compiler output when the right thing is to add
-Wno-deprecated-declarations, fix it.
Fixes: 4ee3c4da8b1b9c22 ("perf scripting python: Do not build fail on deprecation warnings") Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
A clang 15 build reveal several unused-but-set variables, removing the
'foo' variable in tests/mmap-basic.o object to address one of those
cases.
Signed-off-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lore.kernel.org/lkml/20220929140514.226807-2-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Jiri Olsa [Thu, 29 Sep 2022 08:13:47 +0000 (10:13 +0200)]
perf parse-events: Ignore clang 15 warning about variable set but unused in bison produced code
clang 15 now warns:
46 65.20 fedora:rawhide : FAIL clang version 15.0.0 (Fedora 15.0.0-3.fc38)
util/parse-events-bison.c:1401:9: error: variable 'parse_events_nerrs' set but not used [-Werror,-Wunused-but-set-variable]
int yynerrs = 0;
^
#define yynerrs parse_events_nerrs
^
1 error generated.
make[3]: *** [/git/perf-6.0.0-rc7/tools/build/Makefile.build:139: util] Error 2
Just ignore one more compiler warning for the bison generated C code.
Committer notes:
Older clangs don't know about -Wunused-but-set-variable, so we need to
add -Wno-unknown-warning-option to avoid this:
37 44.92 fedora:32 : FAIL clang version 10.0.1 (Fedora 10.0.1-3.fc32)
error: unknown warning option '-Wno-unused-but-set-variable'; did you mean '-Wno-unused-const-variable'? [-Werror,-Wunknown-warning-option]
make[3]: *** [/git/perf-6.0.0-rc7/tools/build/Makefile.build:139: util] Error 2
Reported-by: Arnaldo Carvalho de Melo <acme@kernel.org> Signed-off-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@kernel.org> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/lkml/20220929140514.226807-1-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Merge tag 'net-6.0-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Paolo Abeni:
"Including fixes from wifi and can.
Current release - regressions:
- phy: don't WARN for PHY_UP state in mdio_bus_phy_resume()
- wifi: fix locking in mac80211 mlme
- eth:
- revert "net: mvpp2: debugfs: fix memory leak when using debugfs_lookup()"
- mlxbf_gige: fix an IS_ERR() vs NULL bug in mlxbf_gige_mdio_probe
Previous releases - regressions:
- wifi: fix regression with non-QoS drivers
Previous releases - always broken:
- mptcp: fix unreleased socket in accept queue
- wifi:
- don't start TX with fq->lock to fix deadlock
- fix memory corruption in minstrel_ht_update_rates()
- eth:
- macb: fix ZynqMP SGMII non-wakeup source resume failure
- mt7531: only do PLL once after the reset
- usbnet: fix memory leak in usbnet_disconnect()
Misc:
- usb: qmi_wwan: add new usb-id for Dell branded EM7455"
* tag 'net-6.0-rc8' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (30 commits)
mptcp: fix unreleased socket in accept queue
mptcp: factor out __mptcp_close() without socket lock
net: ethernet: mtk_eth_soc: fix mask of RX_DMA_GET_SPORT{,_V2}
net: mscc: ocelot: fix tagged VLAN refusal while under a VLAN-unaware bridge
can: c_can: don't cache TX messages for C_CAN cores
ice: xsk: drop power of 2 ring size restriction for AF_XDP
ice: xsk: change batched Tx descriptor cleaning
net: usb: qmi_wwan: Add new usb-id for Dell branded EM7455
selftests: Fix the if conditions of in test_extra_filter()
net: phy: Don't WARN for PHY_UP state in mdio_bus_phy_resume()
net: stmmac: power up/down serdes in stmmac_open/release
wifi: mac80211: mlme: Fix double unlock on assoc success handling
wifi: mac80211: mlme: Fix missing unlock on beacon RX
wifi: mac80211: fix memory corruption in minstrel_ht_update_rates()
wifi: mac80211: fix regression with non-QoS drivers
wifi: mac80211: ensure vif queues are operational after start
wifi: mac80211: don't start TX with fq->lock to fix deadlock
wifi: cfg80211: fix MCS divisor value
net: hippi: Add missing pci_disable_device() in rr_init_one()
net/mlxbf_gige: Fix an IS_ERR() vs NULL bug in mlxbf_gige_mdio_probe
...
Merge tag 'nvme-6.0-2022-09-29' of git://git.infradead.org/nvme into block-6.0
Pull NVMe fixes from Christoph:
"nvme fixes for Linux 6.1
- fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices
(Michael Kelley)
- disable Write Zeroes on Phison E3C/E4C (Tina Hsu)"
* tag 'nvme-6.0-2022-09-29' of git://git.infradead.org/nvme:
nvme-pci: disable Write Zeroes on Phison E3C/E4C
nvme: Fix IOC_PR_CLEAR and IOC_PR_RELEASE ioctls for nvme devices
Leo Li [Mon, 26 Sep 2022 21:53:55 +0000 (17:53 -0400)]
drm/amd/display: Prevent OTG shutdown during PSR SU
[Why]
Enabling Z10 optimizations allows DMUB to disable the OTG during PSR
link-off. This theoretically saves power by putting more of the display
hardware to sleep. However, we observe that with PSR SU, it causes
visual artifacts, higher power usage, and potential system hang.
This is partly due to an odd behavior with the VStartup interrupt used
to signal DRM vblank events. If the OTG is toggled on/off during a PSR
link on/off cycle, the vstartup interrupt fires twice in quick
succession. This generates incorrectly timed vblank events.
Additionally, it can cause cursor updates to generate visual artifacts.
Note that this is not observed with PSR1 since PSR is fully disabled
when there are vblank event requestors. Cursor updates are also
artifact-free, likely because there are no selectively-updated (SU)
frames that can generate artifacts.
[How]
A potential solution is to disable z10 idle optimizations only when fast
updates (flips & cursor updates) are committed. A mechanism to do so
would require some thoughtful design. Let's just disable idle
optimizations for PSR2 for now.
Fixes: 7cc191ee7621 ("drm/amd/display: Implement MPO PSR SU") Reported-by: August Wikerfors <git@augustwikerfors.se> Link: https://lore.kernel.org/r/c1f8886a-5624-8f49-31b1-e42b6d20dcf5@augustwikerfors.se/ Tested-by: August Wikerfors <git@augustwikerfors.se> Reviewed-by: Harry Wentland <harry.wentland@amd.com> Signed-off-by: Leo Li <sunpeng.li@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
perf tests record: Fail the test if the 'errs' counter is not zero
We were just checking for the 'err' variable, when we should really see
if there was some of the many checked errors that don't stop the test
right away.
Detected with clang 15.0.0:
44 75.23 fedora:37 : FAIL clang version 15.0.0 (Fedora 15.0.0-2.fc37)
tests/perf-record.c:68:16: error: variable 'errs' set but not used [-Werror,-Wunused-but-set-variable]
int err = -1, errs = 0, i, wakeups = 0;
^
1 error generated.
The patch introducing this 'perf test' entry had that check:
Fixes: 2cf88f4614c996e5 ("perf test: Use skip in PERF_RECORD_*") Acked-by: Ian Rogers <irogers@google.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lore.kernel.org/lkml/YzR0n5QhsH9VyYB0@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf test: Fix test case 87 ("perf record tests") for hybrid systems
The test case 87 ("perf record tests") failed on hybrid systems,the event
"cpu/br_inst_retired.near_call/p" is only for non-hybrid system. Correct
the test event to support both non-hybrid and hybrid systems.
Before:
# ./perf test 87
87: perf record tests : FAILED!
After:
# ./perf test 87
87: perf record tests : Ok
Fixes: 24f378e66021f559 ("perf test: Add basic perf record tests") Reviewed-by: Kan Liang <kan.liang@linux.intel.com> Signed-off-by: Xing Zhengjun <zhengjun.xing@linux.intel.com> Acked-by: Ian Rogers <irogers@google.com> Cc: Alexander Shishkin <alexander.shishkin@intel.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/r/20220927051513.3768717-1-zhengjun.xing@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Merge tag 'ata-6.0-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata
Pull ATA fixes from Damien Le Moal:
"Three late patches to fix problems discovered recently:
- Add a horkage to disable link power management by default for the
Pioneer BDR-207M and BDR-205 DVD drives (from Niklas)
- Two patches to fix setting the maximum queue depth of libsas owned
ATA devices (from me)"
* tag 'ata-6.0-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata:
ata: libata-sata: Fix device queue depth control
ata: libata-scsi: Fix initialization of device queue depth
libata: add ATA_HORKAGE_NOLPM for Pioneer BDR-207M and BDR-205
Merge tag 'loongarch-fixes-6.0-3' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson
Pull LoongArch fixes from Huacai Chen:
"Some trivial fixes and cleanup"
* tag 'loongarch-fixes-6.0-3' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:
LoongArch: Clean up loongson3_smp_ops declaration
LoongArch: Fix and cleanup csr_era handling in do_ri()
LoongArch: Align the address of kernel_entry to 4KB
LoongArch: Fix and cleanup csr_era handling in do_ri()
We don't emulate reserved instructions and just send a signal to the
current process now. So we don't need to call compute_return_era() to
add 4 (point to the next instruction) to csr_era in pt_regs. RA/ERA's
backup/restore is cleaned up as well.
Signed-off-by: Jun Yi <yijun@loongson.cn> Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
The mptcp socket and its subflow sockets in accept queue can't be
released after the process exit.
While the release of a mptcp socket in listening state, the
corresponding tcp socket will be released too. Meanwhile, the tcp
socket in the unaccept queue will be released too. However, only init
subflow is in the unaccept queue, and the joined subflow is not in the
unaccept queue, which makes the joined subflow won't be released, and
therefore the corresponding unaccepted mptcp socket will not be released
to.
This can be reproduced easily with following steps:
1. create 2 namespace and veth:
$ ip netns add mptcp-client
$ ip netns add mptcp-server
$ sysctl -w net.ipv4.conf.all.rp_filter=0
$ ip netns exec mptcp-client sysctl -w net.mptcp.enabled=1
$ ip netns exec mptcp-server sysctl -w net.mptcp.enabled=1
$ ip link add red-client netns mptcp-client type veth peer red-server \
netns mptcp-server
$ ip -n mptcp-server address add 10.0.0.1/24 dev red-server
$ ip -n mptcp-server address add 192.168.0.1/24 dev red-server
$ ip -n mptcp-client address add 10.0.0.2/24 dev red-client
$ ip -n mptcp-client address add 192.168.0.2/24 dev red-client
$ ip -n mptcp-server link set red-server up
$ ip -n mptcp-client link set red-client up
2. configure the endpoint and limit for client and server:
$ ip -n mptcp-server mptcp endpoint flush
$ ip -n mptcp-server mptcp limits set subflow 2 add_addr_accepted 2
$ ip -n mptcp-client mptcp endpoint flush
$ ip -n mptcp-client mptcp limits set subflow 2 add_addr_accepted 2
$ ip -n mptcp-client mptcp endpoint add 192.168.0.2 dev red-client id \
1 subflow
3. listen and accept on a port, such as 9999. The nc command we used
here is modified, which makes it use mptcp protocol by default.
$ ip netns exec mptcp-server nc -l -k -p 9999
4. open another *two* terminal and use each of them to connect to the
server with the following command:
$ ip netns exec mptcp-client nc 10.0.0.1 9999
Input something after connect to trigger the connection of the second
subflow. So that there are two established mptcp connections, with the
second one still unaccepted.
5. exit all the nc command, and check the tcp socket in server namespace.
And you will find that there is one tcp socket in CLOSE_WAIT state
and can't release forever.
Fix this by closing all of the unaccepted mptcp socket in
mptcp_subflow_queue_clean() with __mptcp_close().
Now, we can ensure that all unaccepted mptcp sockets will be cleaned by
__mptcp_close() before they are released, so mptcp_sock_destruct(), which
is used to clean the unaccepted mptcp socket, is not needed anymore.
The selftests for mptcp is ran for this commit, and no new failures.
Fixes: f296234c98a8 ("mptcp: Add handling of incoming MP_JOIN requests") Fixes: 6aeed9045071 ("mptcp: fix race on unaccepted mptcp sockets") Cc: stable@vger.kernel.org Reviewed-by: Jiang Biao <benbjiang@tencent.com> Reviewed-by: Mengen Sun <mengensun@tencent.com> Acked-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Menglong Dong <imagedong@tencent.com> Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
mptcp: factor out __mptcp_close() without socket lock
Factor out __mptcp_close() from mptcp_close(). The caller of
__mptcp_close() should hold the socket lock, and cancel mptcp work when
__mptcp_close() returns true.
This function will be used in the next commit.
Fixes: f296234c98a8 ("mptcp: Add handling of incoming MP_JOIN requests") Fixes: 6aeed9045071 ("mptcp: fix race on unaccepted mptcp sockets") Cc: stable@vger.kernel.org Reviewed-by: Jiang Biao <benbjiang@tencent.com> Reviewed-by: Mengen Sun <mengensun@tencent.com> Acked-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Menglong Dong <imagedong@tencent.com> Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Jakub Kicinski [Thu, 29 Sep 2022 02:04:43 +0000 (19:04 -0700)]
Merge branch '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Tony Nguyen says:
====================
ice: xsk: ZC changes
Maciej Fijalkowski says:
This set consists of two fixes to issues that were either pointed out on
indirectly (John was reviewing AF_XDP selftests that were testing ice's
ZC support) mailing list or were directly reported by customers.
First patch allows user space to see done descriptor in CQ even after a
single frame being transmitted and second patch removes the need for
having HW rings sized to power of 2 number of descriptors when used
against AF_XDP.
I also forgot to mention that due to the current Tx cleaning algorithm,
4k HW ring was broken and these two patches bring it back to life, so we
kill two birds with one stone.
* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
ice: xsk: drop power of 2 ring size restriction for AF_XDP
ice: xsk: change batched Tx descriptor cleaning
====================
Vladimir Oltean [Tue, 27 Sep 2022 12:20:42 +0000 (15:20 +0300)]
net: mscc: ocelot: fix tagged VLAN refusal while under a VLAN-unaware bridge
Currently the following set of commands fails:
$ ip link add br0 type bridge # vlan_filtering 0
$ ip link set swp0 master br0
$ bridge vlan
port vlan-id
swp0 1 PVID Egress Untagged
$ bridge vlan add dev swp0 vid 10
Error: mscc_ocelot_switch_lib: Port with more than one egress-untagged VLAN cannot have egress-tagged VLANs.
Dumping ocelot->vlans, one can see that the 2 egress-untagged VLANs on swp0 are
vid 1 (the bridge PVID) and vid 4094, a PVID used privately by the driver for
VLAN-unaware bridging. So this is why bridge vid 10 is refused, despite
'bridge vlan' showing a single egress untagged VLAN.
As mentioned in the comment added, having this private VLAN does not impose
restrictions to the hardware configuration, yet it is a bookkeeping problem.
There are 2 possible solutions.
One is to make the functions that operate on VLAN-unaware pvids:
- ocelot_add_vlan_unaware_pvid()
- ocelot_del_vlan_unaware_pvid()
- ocelot_port_setup_dsa_8021q_cpu()
- ocelot_port_teardown_dsa_8021q_cpu()
call something different than ocelot_vlan_member_(add|del)(), the latter being
the real problem, because it allocates a struct ocelot_bridge_vlan *vlan which
it adds to ocelot->vlans. We don't really *need* the private VLANs in
ocelot->vlans, it's just that we have the extra convenience of having the
vlan->portmask cached in software (whereas without these structures, we'd have
to create a raw ocelot_vlant_rmw_mask() procedure which reads back the current
port mask from hardware).
The other solution is to filter out the private VLANs from
ocelot_port_num_untagged_vlans(), since they aren't what callers care about.
We only need to do this to the mentioned function and not to
ocelot_port_num_tagged_vlans(), because private VLANs are never egress-tagged.
Nothing else seems to be broken in either solution, but the first one requires
more rework which will conflict with the net-next change 36a0bf443585 ("net:
mscc: ocelot: set up tag_8021q CPU ports independent of user port affinity"),
and I'd like to avoid that. So go with the other one.
Peng Fan [Tue, 30 Aug 2022 03:31:31 +0000 (11:31 +0800)]
clk: imx93: drop of_match_ptr
There is build warning when CONFIG_OF is not selected.
>> drivers/clk/imx/clk-imx93.c:324:34: warning: 'imx93_clk_of_match'
>> defined but not used [-Wunused-const-variable=]
324 | static const struct of_device_id imx93_clk_of_match[] = {
| ^~~~~~~~~~~~~~~~~~
The driver only support DT table, no sense to use of_match_ptr.
Fixes: 24defbe194b6 ("clk: imx: add i.MX93 clk") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Peng Fan <peng.fan@nxp.com> Link: https://lore.kernel.org/r/20220830033137.4149542-3-peng.fan@oss.nxp.com Reviewed-by: Abel Vesa <abel.vesa@linaro.org> Signed-off-by: Stephen Boyd <sboyd@kernel.org>
clk: iproc: Do not rely on node name for correct PLL setup
After commit 31fd9b79dc58 ("ARM: dts: BCM5301X: update CRU block
description") a warning from clk-iproc-pll.c was generated due to a
duplicate PLL name as well as the console stopped working. Upon closer
inspection it became clear that iproc_pll_clk_setup() used the Device
Tree node unit name as an unique identifier as well as a parent name to
parent all clocks under the PLL.
BCM5301X was the first platform on which that got noticed because of the
DT node unit name renaming but the same assumptions hold true for any
user of the iproc_pll_clk_setup() function.
The first 'clock-output-names' property is always guaranteed to be
unique as well as providing the actual desired PLL clock name, so we
utilize that to register the PLL and as a parent name of all children
clock.
Stefan reports that there are issues with the level triggered
notification. Since we're late in the cycle, and it was introduced for
the 6.0 release, just disable it at prep time and we can bring this
back when Samba is happy with it.
Reported-by: Stefan Metzmacher <metze@samba.org> Reviewed-by: Stefan Metzmacher <metze@samba.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
In commit 4167ac8a657e ("clk: sunxi-ng: sun50i: h6: Modify GPU clock
configuration to support DFS") divider M0 was forced to be 1 in order to
support DFS. However, that left N as it is, at high value of 36. On
boards without devfreq enabled (all of them in kernel 6.0), this
effectively sets GPU frequency to 864 MHz. This is about 100 MHz above
maximum supported frequency.
In order to fix this, let's set N to 18 (register value 17). That way
default frequency of 432 MHz is preserved.
Ashutosh Dixit [Wed, 28 Sep 2022 19:02:12 +0000 (12:02 -0700)]
drm/i915/gt: Perf_limit_reasons are only available for Gen11+
Register GT0_PERF_LIMIT_REASONS (0x1381a8) is available only for
Gen11+. Therefore ensure perf_limit_reasons sysfs files are created only
for Gen11+. Otherwise on Gen < 5 accessing these files results in the
following oops:
<1> [88.829420] BUG: unable to handle page fault for address: ffffc90000bb81a8
<1> [88.829438] #PF: supervisor read access in kernel mode
<1> [88.829447] #PF: error_code(0x0000) - not-present page
This patch is a backport of the drm-tip commit 0d2d201095e9
("drm/i915: Perf_limit_reasons are only available for Gen11+") to
drm-intel-fixes. The backport is not identical to the original, it only
includes the sysfs portions of if. The debugfs portion is not available
in drm-intel-fixes so has not been backported.
Han Xu [Thu, 15 Sep 2022 15:09:59 +0000 (10:09 -0500)]
clk: imx: imx6sx: remove the SET_RATE_PARENT flag for QSPI clocks
There is no dedicate parent clock for QSPI so SET_RATE_PARENT flag
should not be used. For instance, the default parent clock for QSPI is
pll2_bus, which is also the parent clock for quite a few modules, such
as MMDC, once GPMI NAND set clock rate for EDO5 mode can cause system
hang due to pll2_bus rate changed.
Fixes: f1541e15e38e ("clk: imx6sx: Switch to clk_hw based API") Signed-off-by: Han Xu <han.xu@nxp.com> Link: https://lore.kernel.org/r/20220915150959.3646702-1-han.xu@nxp.com Tested-by: Fabio Estevam <festevam@denx.de> Reviewed-by: Abel Vesa <abel.vesa@linaro.org> Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Merge tag 'irq_urgent_for_v6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull more irqchip fixes from Borislav Petkov:
"More irqchip fixes for 6.0 from Marc Zyngier. Stuff got left hanging
due to the whole Plumbers and vacations commotion.
- A couple of configuration fixes for the recently merged Loongarch
drivers
- A fix to avoid dynamic allocation of a cpumask which was causing
issues with PREEMPT_RT and the GICv3 ITS
- A tightening of an error check in the stm32 exti driver"
* tag 'irq_urgent_for_v6.0' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip/loongson-pch-lpc: Add dependence on LoongArch
irqchip: Select downstream irqchip drivers for LoongArch CPU
irqchip/gic-v3-its: Remove cpumask_var_t allocation
irqchip/stm32-exti: Remove check on always false condition
Merge tag 'mmc-v6.0-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc
Pull MMC fixes from Ulf Hansson:
"A couple of MMC fixes. This time there is also a fix for the ARM SCMI
firmware driver, which has been acked by Sudeep Holla, the maintainer.
MMC core:
- Terminate infinite loop in SD-UHS voltage switch
MMC host:
- hsq: Fix kernel crash in the recovery path
- moxart: Fix bus width configurations
- sdhci: Fix kernel panic for cqe irq
ARM_SCMI:
- Fixup clock management by reverting 'firmware: arm_scmi: Add clock
management to the SCMI power domain'"
* tag 'mmc-v6.0-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/ulfh/mmc:
mmc: hsq: Fix data stomping during mmc recovery
Revert "firmware: arm_scmi: Add clock management to the SCMI power domain"
mmc: core: Terminate infinite loop in SD-UHS voltage switch
mmc: moxart: fix 4-bit bus width and remove 8-bit bus width
mmc: sdhci: Fix host->cmd is null
Al Viro [Mon, 26 Sep 2022 15:59:14 +0000 (11:59 -0400)]
[coredump] don't use __kernel_write() on kmap_local_page()
passing kmap_local_page() result to __kernel_write() is unsafe -
random ->write_iter() might (and 9p one does) get unhappy when
passed ITER_KVEC with pointer that came from kmap_local_page().
Fix by providing a variant of __kernel_write() that takes an iov_iter
from caller (__kernel_write() becomes a trivial wrapper) and adding
dump_emit_page() that parallels dump_emit(), except that instead of
__kernel_write() it uses __kernel_write_iter() with ITER_BVEC source.
Fixes: 3159ed57792b "fs/coredump: use kmap_local_page()" Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Stephen Boyd [Wed, 28 Sep 2022 17:52:14 +0000 (10:52 -0700)]
Merge tag 'clk-microchip-fixes-6.0' of https://git.kernel.org/pub/scm/linux/kernel/git/at91/linux into clk-fixes
Pull Microchip clock fixes for Polarfire SoCs:
- fix panic at boot in clock initialization when building with clang-15
- make RTC's AHB clock critical as rtc will stop if the AHB
interface clock is disabled
* tag 'clk-microchip-fixes-6.0' of https://git.kernel.org/pub/scm/linux/kernel/git/at91/linux:
clk: microchip: mpfs: make the rtc's ahb clock critical
clk: microchip: mpfs: fix clk_cfg array bounds violation
Jakub Kicinski [Wed, 28 Sep 2022 17:48:38 +0000 (10:48 -0700)]
Merge tag 'linux-can-fixes-for-6.0-20220928' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can
Marc Kleine-Budde says:
====================
pull-request: can 2022-09-28
The patch is by me and targets the c_can driver. It disables an
optimization in the TX path of C_CAN cores which causes problems.
* tag 'linux-can-fixes-for-6.0-20220928' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: c_can: don't cache TX messages for C_CAN cores
====================
Borislav Petkov [Fri, 19 Aug 2022 17:47:44 +0000 (19:47 +0200)]
x86/cacheinfo: Add a cpu_llc_shared_mask() UP variant
On a CONFIG_SMP=n kernel, the LLC shared mask is 0, which prevents
__cache_amd_cpumap_setup() from doing the L3 masks setup, and more
specifically from setting up the shared_cpu_map and shared_cpu_list
files in sysfs, leading to lscpu from util-linux getting confused and
segfaulting.
Add a cpu_llc_shared_mask() UP variant which returns a mask with a
single bit set, i.e., for CPU0.
Anatoly Pugachev reported that the commit 1d0403d20f6c ("net: set proper
memcg for net_init hooks allocations") is somehow causing the sparc64
VMs failed to boot and the VMs boot fine with that patch reverted. So,
revert the patch for now and later we can debug the issue.
Link: https://lore.kernel.org/all/20220918092849.GA10314@u164.east.ru/ Reported-by: Anatoly Pugachev <matorola@gmail.com> Signed-off-by: Shakeel Butt <shakeelb@google.com> Cc: Vasily Averin <vvs@openvz.org> Cc: Jakub Kicinski <kuba@kernel.org> Cc: Michal Koutný <mkoutny@suse.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: cgroups@vger.kernel.org Cc: sparclinux@vger.kernel.org Cc: linux-mm@kvack.org Cc: linux-kernel@vger.kernel.org Tested-by: Anatoly Pugachev <matorola@gmail.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Fixes: 1d0403d20f6c ("net: set proper memcg for net_init hooks allocations") Reviewed-by: Muchun Song <songmuchun@bytedance.com> Acked-by: Roman Gushchin <roman.gushchin@linux.dev> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
perf arm-spe: augment the data source type with neoverse_spe list
When synthesizing event with SPE data source, commit 4e6430cbb1a9("perf
arm-spe: Use SPE data source for neoverse cores") augment the type with
source information by MIDR. However, is_midr_in_range only compares the
first entry in neoverse_spe.
Change is_midr_in_range to is_midr_in_range_list to traverse the
neoverse_spe array so that all neoverse cores synthesize event with data
source packet.
Fixes: 4e6430cbb1a9f1dc ("perf arm-spe: Use SPE data source for neoverse cores") Reviewed-by: Ali Saidi <alisaidi@amazon.com> Reviewed-by: Leo Yan <leo.yan@linaro.org> Signed-off-by: Jing Zhang <renyu.zj@linux.alibaba.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ali Saidi <alisaidi@amazon.com> Cc: German Gomez <german.gomez@arm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: James Clark <james.clark@arm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: John Garry <john.garry@huawei.com> Cc: linux-arm-kernel@lists.infradead.org Cc: Mark Rutland <mark.rutland@arm.com> Cc: Mike Leach <mike.leach@linaro.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Shuai Xue <xueshuai@linux.alibaba.com> Cc: Timothy Hayes <timothy.hayes@arm.com> Cc: Will Deacon <will@kernel.org> Cc: Zhuo Song <zhuo.song@linux.alibaba.com> Link: https://lore.kernel.org/r/1664197396-42672-1-git-send-email-renyu.zj@linux.alibaba.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf tests vmlinux-kallsyms: Update is_ignored_symbol function to match the kernel ignored list
The testcase “vmlinux-kallsyms.c” fails in powerpc.
vmlinux symtab matches kallsyms: FAILED!
This test look at the symbols in the vmlinux DSO and check if we find
all of them in the kallsyms dso.
But from the powerpc logs , observed that the failure happens for:
ERR : 0xc0000000000fe9c8: .Lmfspr_table not on kallsyms
ERR : 0xc0000000001009c8: .Lmtspr_table not on kallsyms
These are labels ( with .L) in the source code and has to be ignored.
Reference code with .Lmtspr_table: arch/powerpc/xmon/spr_access.S
The testcases invokes is_ignored_symbol() function to ignore hidden
symbols in the dso like local symbols. This function is adapted from
is_ignored_symbol() kernel function in code: scripts/kallsyms.c . The
kernel function got some updates which is not reflected in the testcase
function and the new updates also handles ignoring "labels".
Below is the changes that went in the kernel function.
/* Symbol names that begin with the following are ignored.*/
static const char * const ignored_prefixes[] = {
"$", /* local symbols for ARM, MIPS, etc. */
- ".LASANPC", /* s390 kasan local symbols */
+ ".L", /* local labels, .LBB,.Ltmpxxx,.L__unnamed_xx,.LASANPC, etc. */
"__crc_", /* modversions */
"__efistub_", /* arm64 EFI stub namespace */
- "__kvm_nvhe_", /* arm64 non-VHE KVM namespace */
+ "__kvm_nvhe_$", /* arm64 local symbols in non-VHE KVM namespace */
+ "__kvm_nvhe_.L", /* arm64 local symbols in non-VHE KVM namespace */
"__AArch64ADRPThunk_", /* arm64 lld */
"__ARMV5PILongThunk_", /* arm lld */
"__ARMV7PILongThunk_",
This change is part of below commits and will handle the
symbols with “.L”
commit d4c858643263 ("kallsyms: ignore all local labels prefixed by '.L'")
commit 6ccf9cb557bd ("KVM: arm64: Symbolize the nVHE HYP addresses")
Update the testcase function to include the new changes.
Reported-by: Disha Goel <disgoel@linux.vnet.ibm.com> Signed-off-by: Athira Jajeev <atrajeev@linux.vnet.ibm.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kajol Jain <kjain@linux.ibm.com> Cc: Madhavan Srinivasan <maddy@linux.vnet.ibm.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Cc: Nageswara R Sastry <rnsastry@linux.ibm.com> Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220928045218.37322-1-atrajeev@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Damien Le Moal [Sat, 24 Sep 2022 06:18:26 +0000 (15:18 +0900)]
ata: libata-sata: Fix device queue depth control
The function __ata_change_queue_depth() uses the helper
ata_scsi_find_dev() to get the ata_device structure of a scsi device and
set that device maximum queue depth. However, when the ata device is
managed by libsas, ata_scsi_find_dev() returns NULL, turning
__ata_change_queue_depth() into a nop, which prevents the user from
setting the maximum queue depth of ATA devices used with libsas based
HBAs.
Fix this by renaming __ata_change_queue_depth() to
ata_change_queue_depth() and adding a pointer to the ata_device
structure of the target device as argument. This pointer is provided by
ata_scsi_change_queue_depth() using ata_scsi_find_dev() in the case of
a libata managed device and by sas_change_queue_depth() using
sas_to_ata_dev() in the case of a libsas managed ata device.
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Tested-by: John Garry <john.garry@huawei.com>
Damien Le Moal [Sat, 24 Sep 2022 05:44:11 +0000 (14:44 +0900)]
ata: libata-scsi: Fix initialization of device queue depth
For SATA devices supporting NCQ, drivers using libsas first initialize a
scsi device queue depth based on the controller and device capabilities,
leading to the scsi device queue_depth field being 32 (ATA maximum queue
depth) for most setup. However, if libata was loaded using the
force=[ID]]noncq argument, the default queue depth should be set to 1 to
reflect the fact that queuable commands will never be used. This is
consistent with manually setting a device queue depth to 1 through sysfs
as that disables NCQ use for the device.
Fix ata_scsi_dev_config() to honor the noncq parameter by sertting the
device queue depth to 1 for devices that do not have the ATA_DFLAG_NCQ
flag set.
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com> Tested-by: John Garry <john.garry@huawei.com>
can: c_can: don't cache TX messages for C_CAN cores
As Jacob noticed, the optimization introduced in 387da6bc7a82 ("can:
c_can: cache frames to operate as a true FIFO") doesn't properly work
on C_CAN, but on D_CAN IP cores. The exact reasons are still unknown.
For now disable caching if CAN frames in the TX path for C_CAN cores.
git.samba.org / sfrench / cifs-2.6.git / log