netfilter: nf_tables: reject table flag and netdev basechain updates
netdev basechain updates are stored in the transaction object hook list.
When setting on the table dormant flag, it iterates over the existing
hooks in the basechain. Thus, skipping the hooks that are being
added/deleted in this transaction, which leaves hook registration in
inconsistent state.
Reject table flag updates in combination with netdev basechain updates
in the same batch:
- Update table flags and add/delete basechain: Check from basechain update
path if there are pending flag updates for this table.
- add/delete basechain and update table flags: Iterate over the transaction
list to search for basechain updates from the table update path.
In both cases, the batch is rejected. Based on suggestion from Florian Westphal.
Fixes: b9703ed44ffb ("netfilter: nf_tables: support for adding new devices to an existing netdev chain") Fixes: 7d937b107108f ("netfilter: nf_tables: support for deleting devices in an existing netdev chain") Signed-off-by: Pablo Neira Ayuso <[email protected]>
netfilter: nf_tables: reject destroy command to remove basechain hooks
Report EOPNOTSUPP if NFT_MSG_DESTROYCHAIN is used to delete hooks in an
existing netdev basechain, thus, only NFT_MSG_DELCHAIN is allowed.
Fixes: 7d937b107108f ("netfilter: nf_tables: support for deleting devices in an existing netdev chain") Signed-off-by: Pablo Neira Ayuso <[email protected]>
Masahiro Yamada [Sat, 23 Mar 2024 11:45:11 +0000 (20:45 +0900)]
modpost: do not make find_tosym() return NULL
As mentioned in commit 397586506c3d ("modpost: Add '.ltext' and
'.ltext.*' to TEXT_SECTIONS"), modpost can result in a segmentation
fault due to a NULL pointer dereference in default_mismatch_handler().
find_tosym() can return the original symbol pointer instead of NULL
if a better one is not found.
This fixes the reported segmentation fault.
Fixes: a23e7584ecf3 ("modpost: unify 'sym' and 'to' in default_mismatch_handler()") Reported-by: Nathan Chancellor <[email protected]> Signed-off-by: Masahiro Yamada <[email protected]>
Masahiro Yamada [Sat, 23 Mar 2024 08:51:01 +0000 (17:51 +0900)]
kconfig: do not reparent the menu inside a choice block
The boolean 'choice' is used to list exclusively selected config
options.
You must not add a dependency between choice members, because such a
dependency would create an invisible entry.
In the following test case, it is impossible to choose 'C'.
[Test Case 1]
choice
prompt "Choose one, but how to choose C?"
config A
bool "A"
config B
bool "B"
config C
bool "C"
depends on A
endchoice
Hence, Kconfig shows the following error message:
Kconfig:1:error: recursive dependency detected!
Kconfig:1: choice <choice> contains symbol C
Kconfig:10: symbol C is part of choice A
Kconfig:4: symbol A is part of choice <choice>
For a resolution refer to Documentation/kbuild/kconfig-language.rst
subsection "Kconfig recursive dependency limitations"
However, Kconfig does not report anything for the following similar code:
[Test Case 2]
choice
prompt "Choose one, but how to choose B?"
config A
bool "A"
config B
bool "B"
depends on A
config C
bool "C"
endchoice
This is because menu_finalize() reparents the menu tree when an entry
depends on the preceding one.
With reparenting, the menu tree:
choice
|- A
|- B
\- C
... will be transformed into the following structure:
choice
|- A
| \- B
\- C
Consequently, Kconfig considers only 'A' and 'C' as choice members.
This behavior is awkward. The second test case should be an error too.
Jakub Kicinski [Wed, 27 Mar 2024 22:39:17 +0000 (15:39 -0700)]
Merge tag 'wireless-2024-03-27' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless
Kalle Valo says:
====================
wireless fixes for v6.9-rc2
The first fixes for v6.9. Ping-Ke Shih now maintains a separate tree
for Realtek drivers, document that in the MAINTAINERS. Plenty of fixes
for both to stack and iwlwifi. Our kunit tests were working only on um
architecture but that's fixed now.
* tag 'wireless-2024-03-27' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless: (21 commits)
MAINTAINERS: wifi: mwifiex: add Francesco as reviewer
kunit: fix wireless test dependencies
wifi: iwlwifi: mvm: include link ID when releasing frames
wifi: iwlwifi: mvm: handle debugfs names more carefully
wifi: iwlwifi: mvm: guard against invalid STA ID on removal
wifi: iwlwifi: read txq->read_ptr under lock
wifi: iwlwifi: fw: don't always use FW dump trig
wifi: iwlwifi: mvm: rfi: fix potential response leaks
wifi: mac80211: correctly set active links upon TTLM
wifi: iwlwifi: mvm: Configure the link mapping for non-MLD FW
wifi: iwlwifi: mvm: consider having one active link
wifi: iwlwifi: mvm: pick the version of SESSION_PROTECTION_NOTIF
wifi: mac80211: fix prep_connection error path
wifi: cfg80211: fix rdev_dump_mpp() arguments order
wifi: iwlwifi: mvm: disable MLO for the time being
wifi: cfg80211: add a flag to disable wireless extensions
wifi: mac80211: fix ieee80211_bss_*_flags kernel-doc
wifi: mac80211: check/clear fast rx for non-4addr sta VLAN changes
wifi: mac80211: fix mlme_link_id_dbg()
MAINTAINERS: wifi: add git tree for Realtek WiFi drivers
...
====================
Linus Torvalds [Wed, 27 Mar 2024 21:53:56 +0000 (14:53 -0700)]
Merge tag '9p-fixes-for-6.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs
Pull 9p fixes from Eric Van Hensbergen:
"Two of these fix syzbot reported issues, and the other fixes a unused
variable in some configurations"
* tag '9p-fixes-for-6.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs:
fs/9p: fix uninitialized values during inode evict
fs/9p: remove redundant pointer v9ses
fs/9p: fix uaf in in v9fs_stat2inode_dotl
Linus Torvalds [Wed, 27 Mar 2024 20:56:41 +0000 (13:56 -0700)]
Merge tag 'for-6.9-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Pull btrfs fixes from David Sterba:
- fix race when reading extent buffer and 'uptodate' status is missed
by one thread (introduced in 6.5)
- do additional validation of devices using major:minor numbers
- zoned mode fixes:
- use zone-aware super block access during scrub
- fix use-after-free during device replace (found by KASAN)
- also delete zones that are 100% unusable to reclaim space
- extent unpinning fixes:
- fix extent map leak after error handling
- print correct range in error message
- error code and message updates
* tag 'for-6.9-rc1-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
btrfs: fix race in read_extent_buffer_pages()
btrfs: return accurate error code on open failure in open_fs_devices()
btrfs: zoned: don't skip block groups with 100% zone unusable
btrfs: use btrfs_warn() to log message at btrfs_add_extent_mapping()
btrfs: fix message not properly printing interval when adding extent map
btrfs: fix warning messages not printing interval at unpin_extent_range()
btrfs: fix extent map leak in unexpected scenario at unpin_extent_cache()
btrfs: validate device maj:min during open
btrfs: zoned: fix use-after-free in do_zone_finish()
btrfs: zoned: use zone aware sb location for scrub
Linus Torvalds [Wed, 27 Mar 2024 20:30:48 +0000 (13:30 -0700)]
Merge tag 'mm-hotfixes-stable-2024-03-27-11-25' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Pull misc fixes from Andrew Morton:
"Various hotfixes. About half are cc:stable and the remainder address
post-6.8 issues or aren't considered suitable for backporting.
zswap figures prominently in the post-6.8 issues - folloup against the
large amount of changes we have just made to that code.
Apart from that, all over the map"
* tag 'mm-hotfixes-stable-2024-03-27-11-25' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (21 commits)
crash: use macro to add crashk_res into iomem early for specific arch
mm: zswap: fix data loss on SWP_SYNCHRONOUS_IO devices
selftests/mm: fix ARM related issue with fork after pthread_create
hexagon: vmlinux.lds.S: handle attributes section
userfaultfd: fix deadlock warning when locking src and dst VMAs
tmpfs: fix race on handling dquot rbtree
selftests/mm: sigbus-wp test requires UFFD_FEATURE_WP_HUGETLBFS_SHMEM
mm: zswap: fix writeback shinker GFP_NOIO/GFP_NOFS recursion
ARM: prctl: reject PR_SET_MDWE on pre-ARMv6
prctl: generalize PR_SET_MDWE support check to be per-arch
MAINTAINERS: remove incorrect M: tag for [email protected]
mm: zswap: fix kernel BUG in sg_init_one
selftests: mm: restore settings from only parent process
tools/Makefile: remove cgroup target
mm: cachestat: fix two shmem bugs
mm: increase folio batch size
mm,page_owner: fix recursion
mailmap: update entry for Leonard Crestez
init: open /initrd.image with O_LARGEFILE
selftests/mm: Fix build with _FORTIFY_SOURCE
...
Jocelyn Falempe [Tue, 12 Mar 2024 09:35:12 +0000 (10:35 +0100)]
drm/vmwgfx: Create debugfs ttm_resource_manager entry only if needed
The driver creates /sys/kernel/debug/dri/0/mob_ttm even when the
corresponding ttm_resource_manager is not allocated.
This leads to a crash when trying to read from this file.
Add a check to create mob_ttm, system_mob_ttm, and gmr_ttm debug file
only when the corresponding ttm_resource_manager is allocated.
Matt Bobrowski [Tue, 26 Mar 2024 19:50:19 +0000 (19:50 +0000)]
bpf: update BPF LSM designated reviewer list
Adding myself in place of both Brendan and Florent as both have since
moved on from working on the BPF LSM and will no longer be devoting
their time to maintaining the BPF LSM.
Chuck Lever [Tue, 26 Mar 2024 21:00:20 +0000 (17:00 -0400)]
NFSD: CREATE_SESSION must never cache NFS4ERR_DELAY replies
There are one or two cases where CREATE_SESSION returns
NFS4ERR_DELAY in order to force the client to wait a bit and try
CREATE_SESSION again. However, after commit e4469c6cc69b ("NFSD: Fix
the NFSv4.1 CREATE_SESSION operation"), NFSD caches that response in
the CREATE_SESSION slot. Thus, when the client resends the
CREATE_SESSION, the server always returns the cached NFS4ERR_DELAY
response rather than actually executing the request and properly
recording its outcome. This blocks the client from making further
progress.
RFC 8881 Section 15.1.1.3 says:
> If NFS4ERR_DELAY is returned on an operation other than SEQUENCE
> that validly appears as the first operation of a request ... [t]he
> request can be retried in full without modification. In this case
> as well, the replier MUST avoid returning a response containing
> NFS4ERR_DELAY as the response to an initial operation of a request
> solely on the basis of its presence in the reply cache.
Neither the original NFSD code nor the discussion in section 18.36.4
refer explicitly to this important requirement, so I missed it.
Note also that not only must the server not cache NFS4ERR_DELAY, but
it has to not advance the CREATE_SESSION slot sequence number so
that it can properly recognize and accept the client's retry.
David Howells [Wed, 27 Mar 2024 14:13:24 +0000 (14:13 +0000)]
cifs: Fix duplicate fscache cookie warnings
fscache emits a lot of duplicate cookie warnings with cifs because the
index key for the fscache cookies does not include everything that the
cifs_find_inode() function does. The latter is used with iget5_locked() to
distinguish between inodes in the local inode cache.
Fix this by adding the creation time and file type to the fscache cookie
key.
Additionally, add a couple of comments to note that if one is changed the
other must be also.
Linus Torvalds [Wed, 27 Mar 2024 17:01:24 +0000 (10:01 -0700)]
Merge tag 'probes-fixes-v6.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace
Pull probes fixlet from Masami Hiramatsu:
- tracing/probes: initialize a 'val' local variable with zero.
This variable is read by FETCH_OP_ST_EDATA in a loop, and is
initialized by FETCH_OP_ARG in the same loop. Since this
initialization is not obvious, smatch warns about it.
Explicitly initializing 'val' with zero fixes this warning.
* tag 'probes-fixes-v6.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
tracing: probes: Fix to zero initialize a local variable
Linus Torvalds [Wed, 27 Mar 2024 16:57:30 +0000 (09:57 -0700)]
Merge tag 'execve-v6.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux
Pull execve fixes from Kees Cook:
- Fix selftests to conform to the TAP output format (Muhammad Usama
Anjum)
- Fix NOMMU linux_binprm::exec pointer in auxv (Max Filippov)
- Replace deprecated strncpy usage (Justin Stitt)
- Replace another /bin/sh instance in selftests
* tag 'execve-v6.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
binfmt: replace deprecated strncpy
exec: Fix NOMMU linux_binprm::exec in transfer_args_to_stack()
selftests/exec: Convert remaining /bin/sh to /bin/bash
selftests/exec: execveat: Improve debug reporting
selftests/exec: recursion-depth: conform test to TAP format output
selftests/exec: load_address: conform test to TAP format output
selftests/exec: binfmt_script: Add the overall result line according to TAP
Andrei Matei [Wed, 27 Mar 2024 02:42:45 +0000 (22:42 -0400)]
bpf: Protect against int overflow for stack access size
This patch re-introduces protection against the size of access to stack
memory being negative; the access size can appear negative as a result
of overflowing its signed int representation. This should not actually
happen, as there are other protections along the way, but we should
protect against it anyway. One code path was missing such protections
(fixed in the previous patch in the series), causing out-of-bounds array
accesses in check_stack_range_initialized(). This patch causes the
verification of a program with such a non-sensical access size to fail.
This check used to exist in a more indirect way, but was inadvertendly
removed in a833a17aeac7.
Andrei Matei [Wed, 27 Mar 2024 02:42:44 +0000 (22:42 -0400)]
bpf: Check bloom filter map value size
This patch adds a missing check to bloom filter creating, rejecting
values above KMALLOC_MAX_SIZE. This brings the bloom map in line with
many other map types.
The lack of this protection can cause kernel crashes for value sizes
that overflow int's. Such a crash was caught by syzkaller. The next
patch adds more guard-rails at a lower level.
Linus Torvalds [Wed, 27 Mar 2024 16:48:47 +0000 (09:48 -0700)]
Fix build errors due to new UIO_MEM_DMA_COHERENT mess
Commit 576882ef5e7f ("uio: introduce UIO_MEM_DMA_COHERENT type")
introduced a new use-case for 'struct uio_mem' where the 'mem' field now
contains a kernel virtual address when 'memtype' is set to
UIO_MEM_DMA_COHERENT.
That in turn causes build errors, because 'mem' is of type
'phys_addr_t', and a virtual address is a pointer type. When the code
just blindly uses cast to mix the two, it caused problems when
phys_addr_t isn't the same size as a pointer - notably on 32-bit
architectures with PHYS_ADDR_T_64BIT.
The proper thing to do would probably be to use a union member, and not
have any casts, and make the 'mem' member be a union of 'mem.physaddr'
and 'mem.vaddr', based on 'memtype'.
This is not that proper thing. This is just fixing the ugly casts to be
even uglier, but at least not cause build errors on 32-bit platforms
with 64-bit physical addresses.
Hari Bathini [Tue, 19 Mar 2024 08:01:52 +0000 (13:31 +0530)]
bpf: fix warning for crash_kexec
With [1], crash dump specific code is moved out of CONFIG_KEXEC_CORE
and placed under CONFIG_CRASH_DUMP, where it is more appropriate.
And since CONFIG_KEXEC & !CONFIG_CRASH_DUMP build option is supported
with that, it led to the below warning:
"WARN: resolve_btfids: unresolved symbol crash_kexec"
Ye Zhang [Thu, 21 Mar 2024 10:21:00 +0000 (18:21 +0800)]
thermal: devfreq_cooling: Fix perf state when calculate dfc res_util
The issue occurs when the devfreq cooling device uses the EM power model
and the get_real_power() callback is provided by the driver.
The EM power table is sorted ascending，can't index the table by cooling
device state，so convert cooling state to performance state by
dfc->max_state - dfc->capped_state.
Johannes Weiner [Thu, 7 Mar 2024 22:07:37 +0000 (17:07 -0500)]
drm/amdgpu: fix deadlock while reading mqd from debugfs
An errant disk backup on my desktop got into debugfs and triggered the
following deadlock scenario in the amdgpu debugfs files. The machine
also hard-resets immediately after those lines are printed (although I
wasn't able to reproduce that part when reading by hand):
amdgpu_debugfs_mqd_read() holds a reservation when it calls
put_user(), which may fault and acquire the mmap_sem. This violates
the established locking order.
Bounce the mqd data through a kernel buffer to get put_user() out of
the illegal section.
Roman Li [Wed, 13 Mar 2024 22:35:13 +0000 (18:35 -0400)]
drm/amd/display: Fix bounds check for dcn35 DcfClocks
[Why]
NumFclkLevelsEnabled is used for DcfClocks bounds check
instead of designated NumDcfClkLevelsEnabled.
That can cause array index out-of-bounds access.
[How]
Use designated variable for dcn35 DcfClocks bounds check.
George Shen [Sat, 17 Feb 2024 00:37:03 +0000 (19:37 -0500)]
drm/amd/display: Remove MPC rate control logic from DCN30 and above
[Why]
MPC flow rate control is not needed for DCN30 and above. Current logic
that uses it can result in underflow for certain edge cases (such as
DSC N422 + ODM combine + 422 left edge pixel).
[How]
Remove MPC flow rate control logic and programming for DCN30 and above.
Taimur Hassan [Thu, 29 Feb 2024 14:52:05 +0000 (09:52 -0500)]
drm/amd/display: Send DTBCLK disable message on first commit
[Why]
Previous patch to allow DTBCLK disable didn't address boot case. Driver
thinks DTBCLK is disabled by default, so we don't send disable message to
PMFW. DTBCLK is then enabled at idle desktop on boot, burning power.
[How]
Set dtbclk_en to true on boot so that disable message is sent during first
commit.
Hamza Mahfooz [Thu, 21 Mar 2024 20:09:21 +0000 (16:09 -0400)]
drm/amd/display: fix IPX enablement
We need to re-enable idle power optimizations after entering PSR. Since,
we get kicked out of idle power optimizations before entering PSR
(entering PSR requires us to write to DCN registers, which isn't allowed
while we are in IPS).
Fixes: a9b1a4f684b3 ("drm/amd/display: Add more checks for exiting idle in DC") Tested-by: Mark Broadworth <[email protected]> Reviewed-by: Nicholas Kazlauskas <[email protected]> Signed-off-by: Hamza Mahfooz <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
Jonathan Kim [Fri, 1 Mar 2024 17:25:17 +0000 (12:25 -0500)]
drm/amdkfd: range check cp bad op exception interrupts
Due to a CP interrupt bug, bad packet garbage exception codes are raised.
Do a range check so that the debugger and runtime do not receive garbage
codes.
Update the user api to guard exception code type checking as well.
Harry Wentland [Tue, 12 Mar 2024 15:21:32 +0000 (11:21 -0400)]
Revert "drm/amd/display: Fix sending VSC (+ colorimetry) packets for DP/eDP displays without PSR"
This causes flicker on a bunch of eDP panels. The info_packet code
also caused regressions on other OSes that we haven't' seen on Linux
yet, but that is likely due to the fact that we haven't had a chance
to test those environments on Linux.
Jakub Kicinski [Mon, 25 Mar 2024 15:56:11 +0000 (08:56 -0700)]
selftests: netdevsim: set test timeout to 10 minutes
The longest running netdevsim test, nexthop.sh, currently takes
5 min to finish. Around 260s to be exact, and 310s on a debug kernel.
The default timeout in selftest is 45sec, so we need an explicit
config. Give ourselves some headroom and use 10min.
Commit under Fixes isn't really to "blame" but prior to that
netdevsim tests weren't integrated with kselftest infra
so blaming the tests themselves doesn't seem right, either.
Fixes: 8ff25dac88f6 ("netdevsim: add Makefile for selftests") Signed-off-by: Jakub Kicinski <[email protected]> Reviewed-by: Simon Horman <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Compilation with CONFIG_GENERIC_FRAMER disabled lead to the following
warnings:
framer.h:184:16: warning: no previous prototype for function 'framer_get' [-Wmissing-prototypes]
184 | struct framer *framer_get(struct device *dev, const char *con_id)
framer.h:184:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
184 | struct framer *framer_get(struct device *dev, const char *con_id)
framer.h:189:6: warning: no previous prototype for function 'framer_put' [-Wmissing-prototypes]
189 | void framer_put(struct device *dev, struct framer *framer)
framer.h:189:1: note: declare 'static' if the function is not intended to be used outside of this translation unit
189 | void framer_put(struct device *dev, struct framer *framer)
Add missing 'static inline' qualifiers for these functions.
Gergo Koteles [Tue, 26 Mar 2024 16:18:48 +0000 (17:18 +0100)]
ALSA: hda/tas2781: remove useless dev_dbg from playback_hook
The debug message "Playback action not supported: action" is not useful,
because the action was previously printed, and the list of supported
actions are intentional.
Remove the debug statement from the default switch case.
Gergo Koteles [Tue, 26 Mar 2024 16:18:45 +0000 (17:18 +0100)]
ALSA: hda/tas2781: remove digital gain kcontrol
The "Speaker Digital Gain" kcontrol controls the TAS2781_DVC_LVL (0x1A)
register. Unfortunately the tas2563 does not have DVC_LVL, but has
INT_MASK0 in 0x1A, which has been misused so far.
Since commit c1947ce61ff4 ("ALSA: hda/realtek: tas2781: enable subwoofer
volume control") the volume of the tas2781 amplifiers can be controlled
by the master volume, so this digital gain kcontrol is not needed.
Arnd Bergmann [Tue, 26 Mar 2024 22:38:07 +0000 (23:38 +0100)]
ALSA: aoa: avoid false-positive format truncation warning
clang warns about what it interprets as a truncated snprintf:
sound/aoa/soundbus/i2sbus/core.c:171:6: error: 'snprintf' will always be truncated; specified size is 6, but format string expands to at least 7 [-Werror,-Wformat-truncation-non-kprintf]
The actual problem here is that it does not understand the special
%pOFn format string and assumes that it is a pointer followed by
the string "OFn", which would indeed not fit.
Slightly increasing the size of the buffer to its natural alignment
avoids the warning, as it is now long enough for the correct and
the incorrect interprations.
Jakub Kicinski [Wed, 27 Mar 2024 03:54:21 +0000 (20:54 -0700)]
Merge branch '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2024-03-25 (ice, ixgbe, igc)
This series contains updates to ice, ixgbe, and igc drivers.
Steven fixes incorrect casting of bitmap type for ice driver.
Jesse fixes memory corruption issue with suspend flow on ice.
Przemek adds GFP_ATOMIC flag to avoid sleeping in IRQ context for ixgbe.
Kurt Kanzenbach removes no longer valid comment on igc.
* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
igc: Remove stale comment about Tx timestamping
ixgbe: avoid sleeping allocation in ixgbe_ipsec_vf_add_sa()
ice: fix memory corruption bug with suspend and rebuild
ice: Refactor FW data type and fix bitmap casting issue
====================
David Thompson [Mon, 25 Mar 2024 18:36:27 +0000 (14:36 -0400)]
mlxbf_gige: call request_irq() after NAPI initialized
The mlxbf_gige driver encounters a NULL pointer exception in
mlxbf_gige_open() when kdump is enabled. The sequence to reproduce
the exception is as follows:
a) enable kdump
b) trigger kdump via "echo c > /proc/sysrq-trigger"
c) kdump kernel executes
d) kdump kernel loads mlxbf_gige module
e) the mlxbf_gige module runs its open() as the
the "oob_net0" interface is brought up
f) mlxbf_gige module will experience an exception
during its open(), something like:
The exception happens because there is a pending RX interrupt before the
call to request_irq(RX IRQ) executes. Then, the RX IRQ handler fires
immediately after this request_irq() completes. The RX IRQ handler runs
"napi_schedule()" before NAPI is fully initialized via "netif_napi_add()"
and "napi_enable()", both which happen later in the open() logic.
The logic in mlxbf_gige_open() must fully initialize NAPI before any calls
to request_irq() execute.
Sabrina Dubroca [Mon, 25 Mar 2024 15:56:48 +0000 (16:56 +0100)]
tls: get psock ref after taking rxlock to avoid leak
At the start of tls_sw_recvmsg, we take a reference on the psock, and
then call tls_rx_reader_lock. If that fails, we return directly
without releasing the reference.
Instead of adding a new label, just take the reference after locking
has succeeded, since we don't need it before.
Sabrina Dubroca [Mon, 25 Mar 2024 15:56:47 +0000 (16:56 +0100)]
selftests: tls: add test with a partially invalid iov
Make sure that we don't return more bytes than we actually received if
the userspace buffer was bogus. We expect to receive at least the rest
of rec1, and possibly some of rec2 (currently, we don't, but that
would be ok).
Sabrina Dubroca [Mon, 25 Mar 2024 15:56:46 +0000 (16:56 +0100)]
tls: adjust recv return with async crypto and failed copy to userspace
process_rx_list may not copy as many bytes as we want to the userspace
buffer, for example in case we hit an EFAULT during the copy. If this
happens, we should only count the bytes that were actually copied,
which may be 0.
Subtracting async_copy_bytes is correct in both peek and !peek cases,
because decrypted == async_copy_bytes + peeked for the peek case: peek
is always !ZC, and we can go through either the sync or async path. In
the async case, we add chunk to both decrypted and
async_copy_bytes. In the sync case, we add chunk to both decrypted and
peeked. I missed that in commit 6caaf104423d ("tls: fix peeking with
sync+async decryption").
Sabrina Dubroca [Mon, 25 Mar 2024 15:56:45 +0000 (16:56 +0100)]
tls: recv: process_rx_list shouldn't use an offset with kvec
Only MSG_PEEK needs to copy from an offset during the final
process_rx_list call, because the bytes we copied at the beginning of
tls_sw_recvmsg were left on the rx_list. In the KVEC case, we removed
data from the rx_list as we were copying it, so there's no need to use
an offset, just like in the normal case.
Yazen Ghannam [Tue, 26 Mar 2024 20:32:52 +0000 (20:32 +0000)]
RAS: Avoid build errors when CONFIG_DEBUG_FS=n
A new helper was introduced for RAS modules to be able to get the RAS
subsystem debugfs root directory. The helper is defined in debugfs.c
which is only built when CONFIG_DEBUG_FS=y.
However, it's possible that the modules would include debugfs support
for optional functionality. One current example is the fmpm module. In
this case, a build error will occur when CONFIG_RAS_FMPM is selected and
CONFIG_DEBUG_FS=n.
Add an inline helper function stub for the CONFIG_DEBUG_FS=n case as the
fmpm module can function without the debugfs functionality too.
Baoquan He [Mon, 25 Mar 2024 01:50:50 +0000 (09:50 +0800)]
crash: use macro to add crashk_res into iomem early for specific arch
There are regression reports[1][2] that crashkernel region on x86_64 can't
be added into iomem tree sometime. This causes the later failure of kdump
loading.
This happened after commit 4a693ce65b18 ("kdump: defer the insertion of
crashkernel resources") was merged.
Even though, these reported issues are proved to be related to other
component, they are just exposed after above commmit applied, I still
would like to keep crashk_res and crashk_low_res being added into iomem
early as before because the early adding has been always there on x86_64
and working very well. For safety of kdump, Let's change it back.
Here, add a macro HAVE_ARCH_ADD_CRASH_RES_TO_IOMEM_EARLY to limit that
only ARCH defining the macro can have the early adding
crashk_res/_low_res into iomem. Then define
HAVE_ARCH_ADD_CRASH_RES_TO_IOMEM_EARLY on x86 to enable it.
Note: In reserve_crashkernel_low(), there's a remnant of crashk_low_res
handling which was mistakenly added back in commit 85fcde402db1 ("kexec:
split crashkernel reservation code out from crash_core.c").
[1]
[PATCH V2] x86/kexec: do not update E820 kexec table for setup_data
https://lore.kernel.org/all/[email protected]/T/#u
[2]
Question about Address Range Validation in Crash Kernel Allocation
https://lore.kernel.org/all/4eeac1f733584855965a2ea62fa4da58@huawei.com/T/#u
Johannes Weiner [Sun, 24 Mar 2024 21:04:47 +0000 (17:04 -0400)]
mm: zswap: fix data loss on SWP_SYNCHRONOUS_IO devices
Zhongkun He reports data corruption when combining zswap with zram.
The issue is the exclusive loads we're doing in zswap. They assume
that all reads are going into the swapcache, which can assume
authoritative ownership of the data and so the zswap copy can go.
However, zram files are marked SWP_SYNCHRONOUS_IO, and faults will try to
bypass the swapcache. This results in an optimistic read of the swap data
into a page that will be dismissed if the fault fails due to races. In
this case, zswap mustn't drop its authoritative copy.
Edward Liaw [Mon, 25 Mar 2024 19:40:52 +0000 (19:40 +0000)]
selftests/mm: fix ARM related issue with fork after pthread_create
Following issue was observed while running the uffd-unit-tests selftest
on ARM devices. On x86_64 no issues were detected:
pthread_create followed by fork caused deadlock in certain cases wherein
fork required some work to be completed by the created thread. Used
synchronization to ensure that created thread's start function has started
before invoking fork.
After the linked LLVM change, the build fails with
CONFIG_LD_ORPHAN_WARN_LEVEL="error", which happens with allmodconfig:
ld.lld: error: vmlinux.a(init/main.o):(.hexagon.attributes) is being placed in '.hexagon.attributes'
Handle the attributes section in a similar manner as arm and riscv by
adding it after the primary ELF_DETAILS grouping in vmlinux.lds.S, which
fixes the error.
Carlos Maiolino [Wed, 20 Mar 2024 12:39:59 +0000 (13:39 +0100)]
tmpfs: fix race on handling dquot rbtree
A syzkaller reproducer found a race while attempting to remove dquot
information from the rb tree.
Fetching the rb_tree root node must also be protected by the
dqopt->dqio_sem, otherwise, giving the right timing, shmem_release_dquot()
will trigger a warning because it couldn't find a node in the tree, when
the real reason was the root node changing before the search starts:
Edward Liaw [Thu, 21 Mar 2024 23:20:21 +0000 (23:20 +0000)]
selftests/mm: sigbus-wp test requires UFFD_FEATURE_WP_HUGETLBFS_SHMEM
The sigbus-wp test requires the UFFD_FEATURE_WP_HUGETLBFS_SHMEM flag for
shmem and hugetlb targets. Otherwise it is not backwards compatible with
kernels <5.19 and fails with EINVAL.
The zswap shrinker resumes the swap_writepage()s that were intercepted
by the zswap store. This will enter the block layer, and may even
enter the filesystem depending on the swap backing file.
Zev Weiss [Tue, 27 Feb 2024 01:35:42 +0000 (17:35 -0800)]
ARM: prctl: reject PR_SET_MDWE on pre-ARMv6
On v5 and lower CPUs we can't provide MDWE protection, so ensure we fail
any attempt to enable it via prctl(PR_SET_MDWE).
Previously such an attempt would misleadingly succeed, leading to any
subsequent mmap(PROT_READ|PROT_WRITE) or execve() failing unconditionally
(the latter somewhat violently via force_fatal_sig(SIGSEGV) due to
READ_IMPLIES_EXEC).
Zev Weiss [Tue, 27 Feb 2024 01:35:41 +0000 (17:35 -0800)]
prctl: generalize PR_SET_MDWE support check to be per-arch
Patch series "ARM: prctl: Reject PR_SET_MDWE where not supported".
I noticed after a recent kernel update that my ARM926 system started
segfaulting on any execve() after calling prctl(PR_SET_MDWE). After some
investigation it appears that ARMv5 is incapable of providing the
appropriate protections for MDWE, since any readable memory is also
implicitly executable.
The prctl_set_mdwe() function already had some special-case logic added
disabling it on PARISC (commit 793838138c15, "prctl: Disable
prctl(PR_SET_MDWE) on parisc"); this patch series (1) generalizes that
check to use an arch_*() function, and (2) adds a corresponding override
for ARM to disable MDWE on pre-ARMv6 CPUs.
With the series applied, prctl(PR_SET_MDWE) is rejected on ARMv5 and
subsequent execve() calls (as well as mmap(PROT_READ|PROT_WRITE)) can
succeed instead of unconditionally failing; on ARMv6 the prctl works as it
did previously.
There exist systems other than PARISC where MDWE may not be feasible to
support; rather than cluttering up the generic code with additional
arch-specific logic let's add a generic function for checking MDWE support
and allow each arch to override it as needed.
The [email protected] mailing list should only be listed under the
L: (List) tag in the MAINTAINERS file. However, it was incorrectly listed
under both L: and M: (Maintainers) tags, which is not accurate. Remove
the M: tag for [email protected] in the MAINTAINERS file to reflect
the correct categorization.
Consequently, we have two choices: either employ kmap_to_page() alongside
sg_set_page(), or resort to copying high memory contents to a temporary
buffer residing in low memory. However, considering the introduction of
the WARN_ON_ONCE in commit ef6e06b2ef870 ("highmem: fix kmap_to_page() for
kmap_local_page() addresses"), which specifically addresses high memory
concerns, it appears that memcpy remains the sole viable option.
selftests: mm: restore settings from only parent process
The atexit() is called from parent process as well as forked processes.
Hence the child restores the settings at exit while the parent is still
executing. Fix this by checking pid of atexit() calling process and only
restore THP number from parent process.
Cong Liu [Fri, 15 Mar 2024 01:22:48 +0000 (09:22 +0800)]
tools/Makefile: remove cgroup target
The tools/cgroup directory no longer contains a Makefile. This patch
updates the top-level tools/Makefile to remove references to building and
installing cgroup components. This change reflects the current structure
of the tools directory and fixes the build failure when building tools in
the top-level directory.
linux/tools$ make cgroup
DESCEND cgroup
make[1]: *** No targets specified and no makefile found. Stop.
make: *** [Makefile:73: cgroup] Error 2
Johannes Weiner [Fri, 15 Mar 2024 09:55:56 +0000 (05:55 -0400)]
mm: cachestat: fix two shmem bugs
When cachestat on shmem races with swapping and invalidation, there
are two possible bugs:
1) A swapin error can have resulted in a poisoned swap entry in the
shmem inode's xarray. Calling get_shadow_from_swap_cache() on it
will result in an out-of-bounds access to swapper_spaces[].
Validate the entry with non_swap_entry() before going further.
2) When we find a valid swap entry in the shmem's inode, the shadow
entry in the swapcache might not exist yet: swap IO is still in
progress and we're before __remove_mapping; swapin, invalidation,
or swapoff have removed the shadow from swapcache after we saw the
shmem swap entry.
This will send a NULL to workingset_test_recent(). The latter
purely operates on pointer bits, so it won't crash - node 0, memcg
ID 0, eviction timestamp 0, etc. are all valid inputs - but it's a
bogus test. In theory that could result in a false "recently
evicted" count.
Such a false positive wouldn't be the end of the world. But for
code clarity and (future) robustness, be explicit about this case.
Bail on get_shadow_from_swap_cache() returning NULL.
On a 104 thread, 2 socket Skylake system, Intel report a 4.7% performance
reduction with will-it-scale page_fault2. This was due to reducing the
size of the batch from 32 to 15. Increasing the folio batch size from 15
to 31 gives a performance increase of 12.5% relative to the original, or
17.2% relative to the reduced performance commit.
The penalty of this commit is an additional 128 bytes of stack usage. Six
folio_batches are also allocated from percpu memory in cpu_fbatches so
that will be an additional 768 bytes of percpu memory (per CPU). Tim Chen
originally submitted a patch like this in 2020:
https://lore.kernel.org/linux-mm/d1cc9f12a8ad6c2a52cb600d93b06b064f2bbc57.1593205965[email protected]/
Oscar Salvador [Fri, 15 Mar 2024 22:26:10 +0000 (23:26 +0100)]
mm,page_owner: fix recursion
Prior to 217b2119b9e2 ("mm,page_owner: implement the tracking of the
stacks count") the only place where page_owner could potentially go into
recursion due to its need of allocating more memory was in save_stack(),
which ends up calling into stackdepot code with the possibility of
allocating memory.
We made sure to guard against that by signaling that the current task was
already in page_owner code, so in case a recursion attempt was made, we
could catch that and return dummy_handle.
After above commit, a new place in page_owner code was introduced where we
could allocate memory, meaning we could go into recursion would we take
that path.
Make sure to signal that we are in page_owner in that codepath as well.
Move the guard code into two helpers {un}set_current_in_page_owner() and
use them prior to calling in the two functions that might allocate memory.
Vitaly Chikunov [Mon, 18 Mar 2024 02:34:44 +0000 (05:34 +0300)]
selftests/mm: Fix build with _FORTIFY_SOURCE
Add missing flags argument to open(2) call with O_CREAT.
Some tests fail to compile if _FORTIFY_SOURCE is defined (to any valid
value) (together with -O), resulting in similar error messages such as:
In file included from /usr/include/fcntl.h:342,
from gup_test.c:1:
In function 'open',
inlined from 'main' at gup_test.c:206:10:
/usr/include/bits/fcntl2.h:50:11: error: call to '__open_missing_mode' declared with attribute error: open with O_CREAT or O_TMPFILE in second argument needs 3 arguments
50 | __open_missing_mode ();
| ^~~~~~~~~~~~~~~~~~~~~~
_FORTIFY_SOURCE is enabled by default in some distributions, so the
tests are not built by default and are skipped.
open(2) man-page warns about missing flags argument: "if it is not
supplied, some arbitrary bytes from the stack will be applied as the
file mode."
Peter Xu [Wed, 13 Mar 2024 21:31:07 +0000 (17:31 -0400)]
mm/memory: fix missing pte marker for !page on pte zaps
Commit 0cf18e839f64 of large folio zap work broke uffd-wp. Now mm's uffd
unit test "wp-unpopulated" will trigger this WARN_ON_ONCE().
The WARN_ON_ONCE() asserts that an VMA cannot be registered with
userfaultfd-wp if it contains a !normal page, but it's actually possible.
One example is an anonymous vma, register with uffd-wp, read anything will
install a zero page. Then when zap on it, this should trigger.
What's more, removing that WARN_ON_ONCE may not be enough either, because
we should also not rely on "whether it's a normal page" to decide whether
pte marker is needed. For example, one can register wr-protect over some
DAX regions to track writes when UFFD_FEATURE_WP_ASYNC enabled, in which
case it can have page==NULL for a devmap but we may want to keep the
marker around.
John Garry [Mon, 25 Mar 2024 08:35:01 +0000 (08:35 +0000)]
block: Make blk_rq_set_mixed_merge() static
Since commit 8e756373d7c8 ("block: Move bio merge related functions into
blk-merge.c"), blk_rq_set_mixed_merge() has only been referenced in
blk-merge.c, so make it static.
Masahiro Yamada [Sun, 4 Feb 2024 09:46:13 +0000 (18:46 +0900)]
cxl: remove CONFIG_CXL_PMU entry in drivers/cxl/Kconfig
Commit 5d7107c72796 ("perf: CXL Performance Monitoring Unit driver")
added the config entries for CXL_PMU in drivers/cxl/Kconfig and
drivers/perf/Kconfig, so it can be toggled from multiple locations:
[1] Device Drivers
-> PCI support
-> CXL (Compute Expres Link) Devices
-> CXL Performance Monitoring Unit
[2] Device Drivers
-> Performance monitor support
-> CXL Performance Monitoring Unit
This complicates things, and nobody else does this.
I kept the one in drivers/perf/Kconfig because CONFIG_CXL_PMU controls
the compilation of drivers/perf/cxl_pmu.c.
Linus Torvalds [Tue, 26 Mar 2024 16:25:57 +0000 (09:25 -0700)]
Merge tag 'printk-for-6.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/printk/linux
Pull printk fix from Petr Mladek:
- Prevent scheduling in an atomic context when printk() takes over the
console flushing duty
* tag 'printk-for-6.9-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/printk/linux:
printk: Update @console_may_schedule in console_trylock_spinning()
Tavian Barnes [Sat, 16 Mar 2024 01:14:29 +0000 (21:14 -0400)]
btrfs: fix race in read_extent_buffer_pages()
There are reports from tree-checker that detects corrupted nodes,
without any obvious pattern so possibly an overwrite in memory.
After some debugging it turns out there's a race when reading an extent
buffer the uptodate status can be missed.
To prevent concurrent reads for the same extent buffer,
read_extent_buffer_pages() performs these checks:
/* (1) */
if (test_bit(EXTENT_BUFFER_UPTODATE, &eb->bflags))
return 0;
/* (2) */
if (test_and_set_bit(EXTENT_BUFFER_READING, &eb->bflags))
goto done;
At this point, it seems safe to start the actual read operation. Once
that completes, end_bbio_meta_read() does
Normally, this is enough to ensure only one read happens, and all other
callers wait for it to finish before returning. Unfortunately, there is
a racey interleaving:
Thread A | Thread B | Thread C
---------+----------+---------
(1) | |
| (1) |
(2) | |
(3) | |
(4) | |
| (2) |
| | (1)
When this happens, thread B kicks of an unnecessary read. Worse, thread
C will see UPTODATE set and return immediately, while the read from
thread B is still in progress. This race could result in tree-checker
errors like this as the extent buffer is concurrently modified:
Anand Jain [Tue, 19 Mar 2024 02:58:18 +0000 (08:28 +0530)]
btrfs: return accurate error code on open failure in open_fs_devices()
When attempting to exclusive open a device which has no exclusive open
permission, such as a physical device associated with the flakey dm
device, the open operation will fail, resulting in a mount failure.
In this particular scenario, we erroneously return -EINVAL instead of the
correct error code provided by the bdev_open_by_path() function, which is
-EBUSY.
Fix this, by returning error code from the bdev_open_by_path() function.
With this correction, the mount error message will align with that of
ext4 and xfs.
btrfs: zoned: don't skip block groups with 100% zone unusable
Commit f4a9f219411f ("btrfs: do not delete unused block group if it may be
used soon") changed the behaviour of deleting unused block-groups on zoned
filesystems. Starting with this commit, we're using
btrfs_space_info_used() to calculate the number of used bytes in a
space_info. But btrfs_space_info_used() also accounts
btrfs_space_info::bytes_zone_unusable as used bytes.
So if a block group is 100% zone_unusable it is skipped from the deletion
step.
In order not to skip fully zone_unusable block-groups, also check if the
block-group has bytes left that can be used on a zoned filesystem.
Filipe Manana [Wed, 13 Mar 2024 17:14:02 +0000 (17:14 +0000)]
btrfs: use btrfs_warn() to log message at btrfs_add_extent_mapping()
At btrfs_add_extent_mapping(), if we failed to merge the extent map, which
is unexpected and theoretically should never happen, we use WARN_ONCE() to
log a message which is not great because we don't get information about
which filesystem it relates to in case we have multiple btrfs filesystems
mounted. So change this to use btrfs_warn() and surround the error check
with WARN_ON() so we always get a useful stack trace and the condition is
flagged as "unlikely" since it's not expected to ever happen.
Filipe Manana [Wed, 13 Mar 2024 13:02:02 +0000 (13:02 +0000)]
btrfs: fix message not properly printing interval when adding extent map
At btrfs_add_extent_mapping(), if we are unable to merge the existing
extent map, we print a warning message that suggests interval ranges in
the form "[X, Y)", where the first element is the inclusive start offset
of a range and the second element is the exclusive end offset. However
we end up printing the length of the ranges instead of the exclusive end
offsets. So fix this by printing the range end offsets.
Filipe Manana [Wed, 13 Mar 2024 12:49:31 +0000 (12:49 +0000)]
btrfs: fix warning messages not printing interval at unpin_extent_range()
At unpin_extent_range() we print warning messages that are supposed to
print an interval in the form "[X, Y)", with the first element being an
inclusive start offset and the second element being the exclusive end
offset of a range. However we end up printing the range's length instead
of the range's exclusive end offset, so fix that to avoid having confusing
and non-sense messages in case we hit one of these unexpected scenarios.
Filipe Manana [Wed, 13 Mar 2024 11:37:31 +0000 (11:37 +0000)]
btrfs: fix extent map leak in unexpected scenario at unpin_extent_cache()
At unpin_extent_cache() if we happen to find an extent map with an
unexpected start offset, we jump to the 'out' label and never release the
reference we added to the extent map through the call to
lookup_extent_mapping(), therefore resulting in a leak. So fix this by
moving the free_extent_map() under the 'out' label.
Anand Jain [Fri, 1 Mar 2024 00:42:13 +0000 (08:42 +0800)]
btrfs: validate device maj:min during open
Boris managed to create a device capable of changing its maj:min without
altering its device path.
Only multi-devices can be scanned. A device that gets scanned and remains
in the btrfs kernel cache might end up with an incorrect maj:min.
Despite the temp-fsid feature patch did not introduce this bug, it could
lead to issues if the above multi-device is converted to a single device
with a stale maj:min. Subsequently, attempting to mount the same device
with the correct maj:min might mistake it for another device with the same
fsid, potentially resulting in wrongly auto-enabling the temp-fsid feature.
To address this, this patch validates the device's maj:min at the time of
device open and updates it if it has changed since the last scan.
btrfs: zoned: fix use-after-free in do_zone_finish()
Shinichiro reported the following use-after-free triggered by the device
replace operation in fstests btrfs/070.
BTRFS info (device nullb1): scrub: finished on devid 1 with status: 0
==================================================================
BUG: KASAN: slab-use-after-free in do_zone_finish+0x91a/0xb90 [btrfs]
Read of size 8 at addr ffff8881543c8060 by task btrfs-cleaner/3494007
The buggy address belongs to the object at ffff8881543c8000
which belongs to the cache kmalloc-1k of size 1024
The buggy address is located 96 bytes inside of
freed 1024-byte region [ffff8881543c8000, ffff8881543c8400)
cleaner_kthread
btrfs_delete_unused_bgs
btrfs_zone_finish
do_zone_finish <-- refers the freed device
The reason for this is that we're using a cached pointer to the chunk_map
from the block group, but on device replace this cached pointer can
contain stale device entries.
The staleness comes from the fact, that btrfs_block_group::physical_map is
not a pointer to a btrfs_chunk_map but a memory copy of it.
Also take the fs_info::dev_replace::rwsem to prevent
btrfs_dev_replace_update_device_in_mapping_tree() from changing the device
underneath us again.
Note: btrfs_dev_replace_update_device_in_mapping_tree() is holding
fs_info::mapping_tree_lock, but as this is a spinning read/write lock we
cannot take it as the call to blkdev_zone_mgmt() requires a memory
allocation which may not sleep.
But btrfs_dev_replace_update_device_in_mapping_tree() is always called with
the fs_info::dev_replace::rwsem held in write mode.
Jian Shen [Mon, 25 Mar 2024 12:43:11 +0000 (20:43 +0800)]
net: hns3: mark unexcuted loopback test result as UNEXECUTED
Currently, loopback test may be skipped when resetting, but the test
result will still show as 'PASS', because the driver doesn't set
ETH_TEST_FL_FAILED flag. Fix it by setting the flag and
initializating the value to UNEXECUTED.
Yonglong Liu [Mon, 25 Mar 2024 12:43:10 +0000 (20:43 +0800)]
net: hns3: fix kernel crash when devlink reload during pf initialization
The devlink reload process will access the hardware resources,
but the register operation is done before the hardware is initialized.
So, processing the devlink reload during initialization may lead to kernel
crash. This patch fixes this by taking devl_lock during initialization.
Kevin Loughlin [Wed, 13 Mar 2024 12:15:46 +0000 (12:15 +0000)]
x86/sev: Skip ROM range scans and validation for SEV-SNP guests
SEV-SNP requires encrypted memory to be validated before access.
Because the ROM memory range is not part of the e820 table, it is not
pre-validated by the BIOS. Therefore, if a SEV-SNP guest kernel wishes
to access this range, the guest must first validate the range.
The current SEV-SNP code does indeed scan the ROM range during early
boot and thus attempts to validate the ROM range in probe_roms().
However, this behavior is neither sufficient nor necessary for the
following reasons:
* With regards to sufficiency, if EFI_CONFIG_TABLES are not enabled and
CONFIG_DMI_SCAN_MACHINE_NON_EFI_FALLBACK is set, the kernel will
attempt to access the memory at SMBIOS_ENTRY_POINT_SCAN_START (which
falls in the ROM range) prior to validation.
For example, Project Oak Stage 0 provides a minimal guest firmware
that currently meets these configuration conditions, meaning guests
booting atop Oak Stage 0 firmware encounter a problematic call chain
during dmi_setup() -> dmi_scan_machine() that results in a crash
during boot if SEV-SNP is enabled.
* With regards to necessity, SEV-SNP guests generally read garbage
(which changes across boots) from the ROM range, meaning these scans
are unnecessary. The guest reads garbage because the legacy ROM range
is unencrypted data but is accessed via an encrypted PMD during early
boot (where the PMD is marked as encrypted due to potentially mapping
actually-encrypted data in other PMD-contained ranges).
In one exceptional case, EISA probing treats the ROM range as
unencrypted data, which is inconsistent with other probing.
Continuing to allow SEV-SNP guests to use garbage and to inconsistently
classify ROM range encryption status can trigger undesirable behavior.
For instance, if garbage bytes appear to be a valid signature, memory
may be unnecessarily reserved for the ROM range. Future code or other
use cases may result in more problematic (arbitrary) behavior that
should be avoided.
While one solution would be to overhaul the early PMD mapping to always
treat the ROM region of the PMD as unencrypted, SEV-SNP guests do not
currently rely on data from the ROM region during early boot (and even
if they did, they would be mostly relying on garbage data anyways).
As a simpler solution, skip the ROM range scans (and the otherwise-
necessary range validation) during SEV-SNP guest early boot. The
potential SEV-SNP guest crash due to lack of ROM range validation is
thus avoided by simply not accessing the ROM range.
In most cases, skip the scans by overriding problematic x86_init
functions during sme_early_init() to SNP-safe variants, which can be
likened to x86_init overrides done for other platforms (ex: Xen); such
overrides also avoid the spread of cc_platform_has() checks throughout
the tree.
In the exceptional EISA case, still use cc_platform_has() for the
simplest change, given (1) checks for guest type (ex: Xen domain status)
are already performed here, and (2) these checks occur in a subsys
initcall instead of an x86_init function.
projects / linux.git / log