bpf: Fix the corner case with may_goto and jump to the 1st insn.
When the following program is processed by the verifier:
L1: may_goto L2
goto L1
L2: w0 = 0
exit
the may_goto insn is first converted to:
L1: r11 = *(u64 *)(r10 -8)
if r11 == 0x0 goto L2
r11 -= 1
*(u64 *)(r10 -8) = r11
goto L1
L2: w0 = 0
exit
then later as the last step the verifier inserts:
*(u64 *)(r10 -8) = BPF_MAX_LOOPS
as the first insn of the program to initialize loop count.
When the first insn happens to be a branch target of some jmp the
bpf_patch_insn_data() logic will produce:
L1: *(u64 *)(r10 -8) = BPF_MAX_LOOPS
r11 = *(u64 *)(r10 -8)
if r11 == 0x0 goto L2
r11 -= 1
*(u64 *)(r10 -8) = r11
goto L1
L2: w0 = 0
exit
because instruction patching adjusts all jmps and calls, but for this
particular corner case it's incorrect and the L1 label should be one
instruction down, like:
*(u64 *)(r10 -8) = BPF_MAX_LOOPS
L1: r11 = *(u64 *)(r10 -8)
if r11 == 0x0 goto L2
r11 -= 1
*(u64 *)(r10 -8) = r11
goto L1
L2: w0 = 0
exit
and that's what this patch is fixing.
After bpf_patch_insn_data() call adjust_jmp_off() to adjust all jmps
that point to newly insert BPF_ST insn to point to insn after.
Note that bpf_patch_insn_data() cannot easily be changed to accommodate
this logic, since jumps that point before or after a sequence of patched
instructions have to be adjusted with the full length of the patch.
Conceptually it's somewhat similar to "insert" of instructions between other
instructions with weird semantics. Like "insert" before 1st insn would require
adjustment of CALL insns to point to newly inserted 1st insn, but not an
adjustment JMP insns that point to 1st, yet still adjusting JMP insns that
cross over 1st insn (point to insn before or insn after), hence use simple
adjust_jmp_off() logic to fix this corner case. Ideally bpf_patch_insn_data()
would have an auxiliary info to say where 'the start of newly inserted patch
is', but it would be too complex for backport.
Matt Bobrowski [Mon, 17 Jun 2024 13:49:10 +0000 (13:49 +0000)]
bpf: Update BPF LSM maintainer list
After catching up with KP recently, we discussed that I will be now be
responsible for co-maintaining the BPF LSM. Adding myself as
designated maintainer of the BPF LSM, and specifying more files in
which the BPF LSM maintenance responsibilities should now extend out
to. This is at the back of all the BPF kfuncs that have been added
recently, which are fundamentally restricted to being used only from
BPF LSM program types.
====================
bpf: Fix missed var_off related to movsx in verifier
Zac reported a verification issue ([1]) where verification unexpectedly succeeded.
This is due to missing proper var_off setting in verifier related to
movsx insn. I found another similar issue as well. This patch set fixed
both problems and added three inline asm tests to test these fixes.
Yonghong Song [Sat, 15 Jun 2024 17:46:32 +0000 (10:46 -0700)]
bpf: Add missed var_off setting in coerce_subreg_to_size_sx()
In coerce_subreg_to_size_sx(), for the case where upper
sign extension bits are the same for smax32 and smin32
values, we missed to setup properly. This is especially
problematic if both smax32 and smin32's sign extension
bits are 1.
The following is a simple example illustrating the inconsistent
verifier states due to missed var_off:
The var_off=(0x80, 0xf) is not correct, and the correct one should
be var_off=(0xffffff80; 0xf) since from insn 3, we know that at
insn 4, the sign extension bits will be 1. This patch fixed this
issue by setting var_off properly.
Yonghong Song [Sat, 15 Jun 2024 17:46:26 +0000 (10:46 -0700)]
bpf: Add missed var_off setting in set_sext32_default_val()
Zac reported a verification failure and Alexei reproduced the issue
with a simple reproducer ([1]). The verification failure is due to missed
setting for var_off.
Note that after insn 1, the var_off for R7 is (0x0; 0x7f). This is not correct
since upper 24 bits of w7 could be 0 or 1. So correct var_off should be
(0x0; 0xffffffff). Missing var_off setting in set_sext32_default_val() caused later
incorrect analysis in zext_32_to_64(dst_reg) and reg_bounds_sync(dst_reg).
To fix the issue, set var_off correctly in set_sext32_default_val(). The correct
reg state after insn 1 becomes:
1: (bc) w7 = (s8)w3 ;
R3_w=scalar(smin=smin32=0,smax=umax=smax32=umax32=255,var_off=(0x0; 0xff))
R7_w=scalar(smin=0,smax=umax=0xffffffff,smin32=-128,smax32=127,var_off=(0x0; 0xffffffff))
and at insn 2, the verifier correctly determines either branch is possible.
Jakub Kicinski [Sat, 15 Jun 2024 00:57:09 +0000 (17:57 -0700)]
Merge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Daniel Borkmann says:
====================
pull-request: bpf 2024-06-14
We've added 8 non-merge commits during the last 2 day(s) which contain
a total of 9 files changed, 92 insertions(+), 11 deletions(-).
The main changes are:
1) Silence a syzkaller splat under CONFIG_DEBUG_NET=y in pskb_pull_reason()
triggered via __bpf_try_make_writable(), from Florian Westphal.
2) Fix removal of kfuncs during linking phase which then throws a kernel
build warning via resolve_btfids about unresolved symbols,
from Tony Ambardar.
3) Fix a UML x86_64 compilation failure from BPF as pcpu_hot symbol
is not available on User Mode Linux, from Maciej Żenczykowski.
4) Fix a register corruption in reg_set_min_max triggering an invariant
violation in BPF verifier, from Daniel Borkmann.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
bpf: Harden __bpf_kfunc tag against linker kfunc removal
compiler_types.h: Define __retain for __attribute__((__retain__))
bpf: Avoid splat in pskb_pull_reason
bpf: fix UML x86_64 compile failure
selftests/bpf: Add test coverage for reg_set_min_max handling
bpf: Reduce stack consumption in check_stack_write_fixed_off
bpf: Fix reg_set_min_max corruption of fake_reg
MAINTAINERS: mailmap: Update Stanislav's email address
====================
Tony Ambardar [Tue, 4 Jun 2024 05:23:16 +0000 (22:23 -0700)]
bpf: Harden __bpf_kfunc tag against linker kfunc removal
BPF kfuncs are often not directly referenced and may be inadvertently
removed by optimization steps during kernel builds, thus the __bpf_kfunc
tag mitigates against this removal by including the __used macro. However,
this macro alone does not prevent removal during linking, and may still
yield build warnings (e.g. on mips64el):
[...]
LD vmlinux
BTFIDS vmlinux
WARN: resolve_btfids: unresolved symbol bpf_verify_pkcs7_signature
WARN: resolve_btfids: unresolved symbol bpf_lookup_user_key
WARN: resolve_btfids: unresolved symbol bpf_lookup_system_key
WARN: resolve_btfids: unresolved symbol bpf_key_put
WARN: resolve_btfids: unresolved symbol bpf_iter_task_next
WARN: resolve_btfids: unresolved symbol bpf_iter_css_task_new
WARN: resolve_btfids: unresolved symbol bpf_get_file_xattr
WARN: resolve_btfids: unresolved symbol bpf_ct_insert_entry
WARN: resolve_btfids: unresolved symbol bpf_cgroup_release
WARN: resolve_btfids: unresolved symbol bpf_cgroup_from_id
WARN: resolve_btfids: unresolved symbol bpf_cgroup_acquire
WARN: resolve_btfids: unresolved symbol bpf_arena_free_pages
NM System.map
SORTTAB vmlinux
OBJCOPY vmlinux.32
[...]
Update the __bpf_kfunc tag to better guard against linker optimization by
including the new __retain compiler macro, which fixes the warnings above.
Verify the __retain macro with readelf by checking object flags for 'R':
$ readelf -Wa kernel/trace/bpf_trace.o
Section Headers:
[Nr] Name Type Address Off Size ES Flg Lk Inf Al
[...]
[178] .text.bpf_key_put PROGBITS 00000000 6420 0050 00 AXR 0 0 8
[...]
Key to Flags:
[...]
R (retain), D (mbind), p (processor specific)
Tony Ambardar [Tue, 4 Jun 2024 05:23:15 +0000 (22:23 -0700)]
compiler_types.h: Define __retain for __attribute__((__retain__))
Some code includes the __used macro to prevent functions and data from
being optimized out. This macro implements __attribute__((__used__)),
which operates at the compiler and IR-level, and so still allows a linker
to remove objects intended to be kept.
Compilers supporting __attribute__((__retain__)) can address this gap by
setting the flag SHF_GNU_RETAIN on the section of a function/variable,
indicating to the linker the object should be retained. This attribute is
available since gcc 11, clang 13, and binutils 2.36.
Provide a __retain macro implementing __attribute__((__retain__)), whose
first user will be the '__bpf_kfunc' tag.
[ Additional remark from discussion:
Why is CONFIG_LTO_CLANG added here? The __used macro permits garbage
collection at section level, so CLANG_LTO_CLANG without
CONFIG_LD_DEAD_CODE_DATA_ELIMINATION should not change final section
dynamics?
The conditional guard was included to ensure consistent behaviour
between __retain and other features forcing split sections. In
particular, the same guard is used in vmlinux.lds.h to merge split
sections where needed. For example, using __retain in LLVM builds
without CONFIG_LTO was failing CI tests on kernel-patches/bpf because
the kernel didn't boot properly. And in further testing, the kernel
had no issues loading BPF kfunc modules with such split sections, so
the module (partial) linking scripts were left alone. ]
* tag 'wireless-2024-06-14' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
wifi: mac80211: fix monitor channel with chanctx emulation
wifi: mac80211: Avoid address calculations via out of bounds array indexing
wifi: mac80211: Recalc offload when monitor stop
wifi: iwlwifi: scan: correctly check if PSC listen period is needed
wifi: iwlwifi: mvm: fix ROC version check
wifi: iwlwifi: mvm: unlock mvm mutex
wifi: cfg80211: wext: add extra SIOCSIWSCAN data check
wifi: cfg80211: wext: set ssids=NULL for passive scans
====================
Ondrej Mosnacek [Fri, 7 Jun 2024 16:07:53 +0000 (18:07 +0200)]
cipso: make cipso_v4_skbuff_delattr() fully remove the CIPSO options
As the comment in this function says, the code currently just clears the
CIPSO part with IPOPT_NOP, rather than removing it completely and
trimming the packet. The other cipso_v4_*_delattr() functions, however,
do the proper removal and also calipso_skbuff_delattr() makes an effort
to remove the CALIPSO options instead of replacing them with padding.
Some routers treat IPv4 packets with anything (even NOPs) in the option
header as a special case and take them through a slower processing path.
Consequently, hardening guides such as STIG recommend to configure such
routers to drop packets with non-empty IP option headers [1][2]. Thus,
users might expect NetLabel to produce packets with minimal padding (or
at least with no padding when no actual options are present).
Implement the proper option removal to address this and to be closer to
what the peer functions do.
Ondrej Mosnacek [Fri, 7 Jun 2024 16:07:52 +0000 (18:07 +0200)]
cipso: fix total option length computation
As evident from the definition of ip_options_get(), the IP option
IPOPT_END is used to pad the IP option data array, not IPOPT_NOP. Yet
the loop that walks the IP options to determine the total IP options
length in cipso_v4_delopt() doesn't take IPOPT_END into account.
Fix it by recognizing the IPOPT_END value as the end of actual options.
Fixes: 014ab19a69c3 ("selinux: Set socket NetLabel based on connection endpoint") Signed-off-by: Ondrej Mosnacek <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Johannes Berg [Wed, 12 Jun 2024 10:23:51 +0000 (12:23 +0200)]
wifi: mac80211: fix monitor channel with chanctx emulation
After the channel context emulation, there were reports that
changing the monitor channel no longer works. This is because
those drivers don't have WANT_MONITOR_VIF, so the setting the
channel always exits out quickly.
Fix this by always allocating the virtual monitor sdata, and
simply not telling the driver about it unless it wanted to.
This way, we have an interface/sdata to bind the chanctx to,
and the emulation can work correctly.
Daniel Borkmann [Thu, 13 Jun 2024 11:53:09 +0000 (13:53 +0200)]
bpf: Reduce stack consumption in check_stack_write_fixed_off
The fake_reg moved into env->fake_reg given it consumes a lot of stack
space (120 bytes). Migrate the fake_reg in check_stack_write_fixed_off()
as well now that we have it.
Daniel Borkmann [Thu, 13 Jun 2024 11:53:08 +0000 (13:53 +0200)]
bpf: Fix reg_set_min_max corruption of fake_reg
Juan reported that after doing some changes to buzzer [0] and implementing
a new fuzzing strategy guided by coverage, they noticed the following in
one of the probes:
What can be seen here is a register invariant violation on line 19. After
the binary-or in line 18, the verifier knows that bit 2 is set but knows
nothing about the rest of the content which was loaded from a map value,
meaning, range is [2,0x7fffffff] with var_off=(0x2; 0x7ffffffd). When in
line 19 the verifier analyzes the branch, it splits the register states
in reg_set_min_max() into the registers of the true branch (true_reg1,
true_reg2) and the registers of the false branch (false_reg1, false_reg2).
Since the test is w6 != 0x7ffffffd, the src_reg is a known constant.
Internally, the verifier creates a "fake" register initialized as scalar
to the value of 0x7ffffffd, and then passes it onto reg_set_min_max(). Now,
for line 19, it is mathematically impossible to take the false branch of
this program, yet the verifier analyzes it. It is impossible because the
second bit of r6 will be set due to the prior or operation and the
constant in the condition has that bit unset (hex(fd) == binary(1111 1101).
When the verifier first analyzes the false / fall-through branch, it will
compute an intersection between the var_off of r6 and of the constant. This
is because the verifier creates a "fake" register initialized to the value
of the constant. The intersection result later refines both registers in
regs_refine_cond_op():
Since the verifier is analyzing the false branch of the conditional jump,
reg1 is equal to false_reg1 and reg2 is equal to false_reg2, i.e. the reg2
is the "fake" register that was meant to hold a constant value. The resulting
var_off of the intersection says that both registers now hold a known value
of var_off=(0x7fffffff, 0x0) or in other words: this operation manages to
make the verifier think that the "constant" value that was passed in the
jump operation now holds a different value.
Normally this would not be an issue since it should not influence the true
branch, however, false_reg2 and true_reg2 are pointers to the same "fake"
register. Meaning, the false branch can influence the results of the true
branch. In line 24, the verifier assumes R6_w=0, but the actual runtime
value in this case is 1. The fix is simply not passing in the same "fake"
register location as inputs to reg_set_min_max(), but instead making a
copy. Moving the fake_reg into the env also reduces stack consumption by
120 bytes. With this, the verifier successfully rejects invalid accesses
from the test program.
- geneve: fix incorrectly setting lengths of inner headers in the
skb, confusing the drivers and causing mangled packets
- sched: initialize noop_qdisc owner to avoid false-positive
recursion detection (recursing on CPU 0), which bubbles up to user
space as a sendmsg() error, while noop_qdisc should silently drop
- netdevsim: fix backwards compatibility in nsim_get_iflink()
Previous releases - always broken:
- netfilter: ipset: fix race between namespace cleanup and gc in the
list:set type"
* tag 'net-6.10-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (35 commits)
bnxt_en: Adjust logging of firmware messages in case of released token in __hwrm_send()
af_unix: Read with MSG_PEEK loops if the first unread byte is OOB
bnxt_en: Cap the size of HWRM_PORT_PHY_QCFG forwarded response
gve: Clear napi->skb before dev_kfree_skb_any()
ionic: fix use after netif_napi_del()
Revert "igc: fix a log entry using uninitialized netdev"
net: bridge: mst: fix suspicious rcu usage in br_mst_set_state
net: bridge: mst: pass vlan group directly to br_mst_vlan_set_state
net/ipv6: Fix the RT cache flush via sysctl using a previous delay
net: stmmac: replace priv->speed with the portTransmitRate from the tc-cbs parameters
gve: ignore nonrelevant GSO type bits when processing TSO headers
net: pse-pd: Use EOPNOTSUPP error code instead of ENOTSUPP
netfilter: Use flowlabel flow key when re-routing mangled packets
netfilter: ipset: Fix race between namespace cleanup and gc in the list:set type
netfilter: nft_inner: validate mandatory meta and payload
tcp: use signed arithmetic in tcp_rtx_probe0_timed_out()
mailmap: map Geliang's new email address
mptcp: pm: update add_addr counters after connect
mptcp: pm: inc RmAddr MIB counter once per RM_ADDR ID
mptcp: ensure snd_una is properly initialized on connect
...
Linus Torvalds [Thu, 13 Jun 2024 18:07:32 +0000 (11:07 -0700)]
Merge tag 'nfs-for-6.10-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Pull NFS client fixes from Trond Myklebust:
"Bugfixes:
- NFSv4.2: Fix a memory leak in nfs4_set_security_label
- NFSv2/v3: abort nfs_atomic_open_v23 if the name is too long.
- NFS: Add appropriate memory barriers to the sillyrename code
- Propagate readlink errors in nfs_symlink_filler
- NFS: don't invalidate dentries on transient errors
- NFS: fix unnecessary synchronous writes in random write workloads
- NFSv4.1: enforce rootpath check when deciding whether or not to trunk
Other:
- Change email address for Trond Myklebust due to email server concerns"
* tag 'nfs-for-6.10-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFS: add barriers when testing for NFS_FSDATA_BLOCKED
SUNRPC: return proper error from gss_wrap_req_priv
NFSv4.1 enforce rootpath check in fs_location query
NFS: abort nfs_atomic_open_v23 if name is too long.
nfs: don't invalidate dentries on transient errors
nfs: Avoid flushing many pages with NFS_FILE_SYNC
nfs: propagate readlink errors in nfs_symlink_filler
MAINTAINERS: Change email address for Trond Myklebust
NFSv4: Fix memory leak in nfs4_set_security_label
Linus Torvalds [Thu, 13 Jun 2024 17:09:29 +0000 (10:09 -0700)]
Merge tag 'fixes-2024-06-13' of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock
Pull memblock fixes from Mike Rapoport:
"Fix validation of NUMA coverage.
memblock_validate_numa_coverage() was checking for a unset node ID
using NUMA_NO_NODE, but x86 used MAX_NUMNODES when no node ID was
specified by buggy firmware.
Update memblock to substitute MAX_NUMNODES with NUMA_NO_NODE in
memblock_set_node() and use NUMA_NO_NODE in x86::numa_init()"
* tag 'fixes-2024-06-13' of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock:
x86/mm/numa: Use NUMA_NO_NODE when calling memblock_set_node()
memblock: make memblock_set_node() also warn about use of MAX_NUMNODES
Wojciech Drewek [Tue, 4 Jun 2024 12:55:14 +0000 (14:55 +0200)]
ice: implement AQ download pkg retry
ice_aqc_opc_download_pkg (0x0C40) AQ sporadically returns error due
to FW issue. Fix this by retrying five times before moving to
Safe Mode. Sleep for 20 ms before retrying. This was tested with the
4.40 firmware.
Fixes: c76488109616 ("ice: Implement Dynamic Device Personalization (DDP) download") Reviewed-by: Michal Swiatkowski <[email protected]> Signed-off-by: Wojciech Drewek <[email protected]> Reviewed-by: Brett Creeley <[email protected]> Reviewed-by: Przemek Kitszel <[email protected]> Tested-by: Pucha Himasekhar Reddy <[email protected]> (A Contingent worker at Intel) Signed-off-by: Tony Nguyen <[email protected]>
Paul Greenwalt [Thu, 30 May 2024 17:06:17 +0000 (13:06 -0400)]
ice: fix 200G link speed message log
Commit 24407a01e57c ("ice: Add 200G speed/phy type use") added support
for 200G PHY speeds, but did not include 200G link speed message
support. As a result the driver incorrectly reports Unknown for 200G
link speed.
Fix this by adding 200G support to ice_print_link_msg().
Fixes: 24407a01e57c ("ice: Add 200G speed/phy type use") Reviewed-by: Michal Swiatkowski <[email protected]> Signed-off-by: Paul Greenwalt <[email protected]> Reviewed-by: Jesse Brandeburg <[email protected]> Tested-by: Pucha Himasekhar Reddy <[email protected]> (A Contingent worker at Intel) Signed-off-by: Tony Nguyen <[email protected]>
En-Wei Wu [Thu, 30 May 2024 14:21:31 +0000 (22:21 +0800)]
ice: avoid IRQ collision to fix init failure on ACPI S3 resume
A bug in https://bugzilla.kernel.org/show_bug.cgi?id=218906 describes
that irdma would break and report hardware initialization failed after
suspend/resume with Intel E810 NIC (tested on 6.9.0-rc5).
The problem is caused due to the collision between the irq numbers
requested in irdma and the irq numbers requested in other drivers
after suspend/resume.
The irq numbers used by irdma are derived from ice's ice_pf->msix_entries
which stores mappings between MSI-X index and Linux interrupt number.
It's supposed to be cleaned up when suspend and rebuilt in resume but
it's not, causing irdma using the old irq numbers stored in the old
ice_pf->msix_entries to request_irq() when resume. And eventually
collide with other drivers.
This patch fixes this problem. On suspend, we call ice_deinit_rdma() to
clean up the ice_pf->msix_entries (and free the MSI-X vectors used by
irdma if we've dynamically allocated them). On resume, we call
ice_init_rdma() to rebuild the ice_pf->msix_entries (and allocate the
MSI-X vectors if we would like to dynamically allocate them).
Aleksandr Mishin [Tue, 11 Jun 2024 08:25:46 +0000 (11:25 +0300)]
bnxt_en: Adjust logging of firmware messages in case of released token in __hwrm_send()
In case of token is released due to token->state == BNXT_HWRM_DEFERRED,
released token (set to NULL) is used in log messages. This issue is
expected to be prevented by HWRM_ERR_CODE_PF_UNAVAILABLE error code. But
this error code is returned by recent firmware. So some firmware may not
return it. This may lead to NULL pointer dereference.
Adjust this issue by adding token pointer check.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Rao Shoaib [Tue, 11 Jun 2024 08:46:39 +0000 (01:46 -0700)]
af_unix: Read with MSG_PEEK loops if the first unread byte is OOB
Read with MSG_PEEK flag loops if the first byte to read is an OOB byte.
commit 22dd70eb2c3d ("af_unix: Don't peek OOB data without MSG_OOB.")
addresses the loop issue but does not address the issue that no data
beyond OOB byte can be read.
Michael Chan [Wed, 12 Jun 2024 23:17:36 +0000 (16:17 -0700)]
bnxt_en: Cap the size of HWRM_PORT_PHY_QCFG forwarded response
Firmware interface 1.10.2.118 has increased the size of
HWRM_PORT_PHY_QCFG response beyond the maximum size that can be
forwarded. When the VF's link state is not the default auto state,
the PF will need to forward the response back to the VF to indicate
the forced state. This regression may cause the VF to fail to
initialize.
Fix it by capping the HWRM_PORT_PHY_QCFG response to the maximum
96 bytes. The SPEEDS2_SUPPORTED flag needs to be cleared because the
new speeds2 fields are beyond the legacy structure. Also modify
bnxt_hwrm_fwd_resp() to print a warning if the message size exceeds 96
bytes to make this failure more obvious.
Ziwei Xiao [Wed, 12 Jun 2024 00:16:54 +0000 (00:16 +0000)]
gve: Clear napi->skb before dev_kfree_skb_any()
gve_rx_free_skb incorrectly leaves napi->skb referencing an skb after it
is freed with dev_kfree_skb_any(). This can result in a subsequent call
to napi_get_frags returning a dangling pointer.
Fix this by clearing napi->skb before the skb is freed.
Taehee Yoo [Wed, 12 Jun 2024 06:04:46 +0000 (06:04 +0000)]
ionic: fix use after netif_napi_del()
When queues are started, netif_napi_add() and napi_enable() are called.
If there are 4 queues and only 3 queues are used for the current
configuration, only 3 queues' napi should be registered and enabled.
The ionic_qcq_enable() checks whether the .poll pointer is not NULL for
enabling only the using queue' napi. Unused queues' napi will not be
registered by netif_napi_add(), so the .poll pointer indicates NULL.
But it couldn't distinguish whether the napi was unregistered or not
because netif_napi_del() doesn't reset the .poll pointer to NULL.
So, ionic_qcq_enable() calls napi_enable() for the queue, which was
unregistered by netif_napi_del().
igc_ptp_init() needs to be called before igc_reset(), otherwise kernel
crash could be observed. Following the corresponding discussion [1] and
[2] revert this commit.
This set fixes a suspicious RCU usage warning triggered by syzbot[1] in
the bridge's MST code. After I converted br_mst_set_state to RCU, I
forgot to update the vlan group dereference helper. Fix it by using
the proper helper, in order to do that we need to pass the vlan group
which is already obtained correctly by the callers for their respective
context. Patch 01 is a requirement for the fix in patch 02.
Note I did consider rcu_dereference_rtnl() but the churn is much bigger
and in every part of the bridge. We can do that as a cleanup in
net-next.
net: bridge: mst: fix suspicious rcu usage in br_mst_set_state
I converted br_mst_set_state to RCU to avoid a vlan use-after-free
but forgot to change the vlan group dereference helper. Switch to vlan
group RCU deref helper to fix the suspicious rcu usage warning.
net: bridge: mst: pass vlan group directly to br_mst_vlan_set_state
Pass the already obtained vlan group pointer to br_mst_vlan_set_state()
instead of dereferencing it again. Each caller has already correctly
dereferenced it for their context. This change is required for the
following suspicious RCU dereference fix. No functional changes
intended.
Petr Pavlu [Fri, 7 Jun 2024 11:28:28 +0000 (13:28 +0200)]
net/ipv6: Fix the RT cache flush via sysctl using a previous delay
The net.ipv6.route.flush system parameter takes a value which specifies
a delay used during the flush operation for aging exception routes. The
written value is however not used in the currently requested flush and
instead utilized only in the next one.
A problem is that ipv6_sysctl_rtcache_flush() first reads the old value
of net->ipv6.sysctl.flush_delay into a local delay variable and then
calls proc_dointvec() which actually updates the sysctl based on the
provided input.
Fix the problem by switching the order of the two operations.
Linus Torvalds [Wed, 12 Jun 2024 23:58:05 +0000 (16:58 -0700)]
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux
Pull ARM and clkdev fixes from Russell King:
- Fix clkdev - erroring out on long strings causes boot failures, so
don't do this. Still warn about the over-sized strings (which will
never match and thus their registration with clkdev is useless)
- Fix for ftrace with frame pointer unwinder with recent GCC changing
the way frames are stacked.
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rmk/linux:
ARM: 9405/1: ftrace: Don't assume stack frames are contiguous in memory
clkdev: don't fail clkdev_alloc() if over-sized
Jakub Kicinski [Wed, 12 Jun 2024 23:28:59 +0000 (16:28 -0700)]
Merge tag 'nf-24-06-11' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Pablo Neira Ayuso says:
====================
Netfilter fixes for net
The following patchset contains Netfilter fixes for net:
Patch #1 fixes insufficient sanitization of netlink attributes for the
inner expression which can trigger nul-pointer dereference,
from Davide Ornaghi.
Patch #2 address a report that there is a race condition between
namespace cleanup and the garbage collection of the list:set
type. This patch resolves this issue with other minor issues
as well, from Jozsef Kadlecsik.
Patch #3 ip6_route_me_harder() ignores flowlabel/dsfield when ip dscp
has been mangled, this unbreaks ip6 dscp set $v,
from Florian Westphal.
All of these patches address issues that are present in several releases.
* tag 'nf-24-06-11' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf:
netfilter: Use flowlabel flow key when re-routing mangled packets
netfilter: ipset: Fix race between namespace cleanup and gc in the list:set type
netfilter: nft_inner: validate mandatory meta and payload
====================
Linus Torvalds [Wed, 12 Jun 2024 22:08:23 +0000 (15:08 -0700)]
Merge tag 'bcachefs-2024-06-12' of https://evilpiepirate.org/git/bcachefs
Pull bcachefs fixes from Kent Overstreet:
- fix kworker explosion, due to calling submit_bio() (which can block)
from a multithreaded workqueue
- fix error handling in btree node scan
- forward compat fix: kill an old debug assert
- key cache shrinker fixes
This is a partial fix for stalls doing multithreaded creates - there
were various O(n^2) issues the key cache shrinker was hitting [1].
There's more work coming here; I'm working on a patch to delete the
key cache lock, which initial testing shows to be a pretty drastic
performance improvement
- assorted syzbot fixes
Link: https://lore.kernel.org/linux-bcachefs/CAGudoHGenxzk0ZqPXXi1_QDbfqQhGHu+wUwzyS6WmfkUZ1HiXA@mail.gmail.com/
* tag 'bcachefs-2024-06-12' of https://evilpiepirate.org/git/bcachefs:
bcachefs: Fix rcu_read_lock() leak in drop_extra_replicas
bcachefs: Add missing bch_inode_info.ei_flags init
bcachefs: Add missing synchronize_srcu_expedited() call when shutting down
bcachefs: Check for invalid bucket from bucket_gen(), gc_bucket()
bcachefs: Replace bucket_valid() asserts in bucket lookup with proper checks
bcachefs: Fix snapshot_create_lock lock ordering
bcachefs: Fix refcount leak in check_fix_ptrs()
bcachefs: Leave a buffer in the btree key cache to avoid lock thrashing
bcachefs: Fix reporting of freed objects from key cache shrinker
bcachefs: set sb->s_shrinker->seeks = 0
bcachefs: increase key cache shrinker batch size
bcachefs: Enable automatic shrinking for rhashtables
bcachefs: fix the display format for show-super
bcachefs: fix stack frame size in fsck.c
bcachefs: Delete incorrect BTREE_ID_NR assertion
bcachefs: Fix incorrect error handling found_btree_node_is_readable()
bcachefs: Split out btree_write_submit_wq
Remi Pommarel [Sat, 18 May 2024 16:07:33 +0000 (18:07 +0200)]
wifi: mac80211: Recalc offload when monitor stop
When a monitor interface is started, ieee80211_recalc_offload() is
called and 802.11 encapsulation offloading support get disabled so
monitor interface could get native wifi frames directly. But when
this interface is stopped there is no need to keep the 802.11
encpasulation offloading off.
This call ieee80211_recalc_offload() when monitor interface is stopped
so 802.11 encapsulation offloading gets re-activated if possible.
Shaul Triebitz [Wed, 5 Jun 2024 11:05:05 +0000 (14:05 +0300)]
wifi: iwlwifi: mvm: fix ROC version check
For using the ROC command, check that the ROC version
is *greater or equal* to 3, rather than *equal* to 3.
The ROC version was added to the TLV starting from
version 3.
Dmitry Antipov [Fri, 31 May 2024 03:20:10 +0000 (06:20 +0300)]
wifi: cfg80211: wext: add extra SIOCSIWSCAN data check
In 'cfg80211_wext_siwscan()', add extra check whether number of
channels passed via 'ioctl(sock, SIOCSIWSCAN, ...)' doesn't exceed
IW_MAX_FREQUENCIES and reject invalid request with -EINVAL otherwise.
Johannes Berg [Tue, 11 Jun 2024 16:58:16 +0000 (18:58 +0200)]
wifi: cfg80211: wext: set ssids=NULL for passive scans
In nl80211, we always set the ssids of a scan request to
NULL when n_ssids==0 (passive scan). Drivers have relied
on this behaviour in the past, so we fixed it in 6 GHz
scan requests as well, and added a warning so we'd have
assurance the API would always be called that way.
syzbot found that wext doesn't ensure that, so we reach
the check and trigger the warning. Fix the wext code to
set the ssids pointer to NULL when there are none.
Xiaolei Wang [Sat, 8 Jun 2024 14:35:24 +0000 (22:35 +0800)]
net: stmmac: replace priv->speed with the portTransmitRate from the tc-cbs parameters
The current cbs parameter depends on speed after uplinking,
which is not needed and will report a configuration error
if the port is not initially connected. The UAPI exposed by
tc-cbs requires userspace to recalculate the send slope anyway,
because the formula depends on port_transmit_rate (see man tc-cbs),
which is not an invariant from tc's perspective. Therefore, we
use offload->sendslope and offload->idleslope to derive the
original port_transmit_rate from the CBS formula.
gve: ignore nonrelevant GSO type bits when processing TSO headers
TSO currently fails when the skb's gso_type field has more than one bit
set.
TSO packets can be passed from userspace using PF_PACKET, TUNTAP and a
few others, using virtio_net_hdr (e.g., PACKET_VNET_HDR). This includes
virtualization, such as QEMU, a real use-case.
The gso_type and gso_size fields as passed from userspace in
virtio_net_hdr are not trusted blindly by the kernel. It adds gso_type
|= SKB_GSO_DODGY to force the packet to enter the software GSO stack
for verification.
This issue might similarly come up when the CWR bit is set in the TCP
header for congestion control, causing the SKB_GSO_TCP_ECN gso_type bit
to be set.
Fixes: a57e5de476be ("gve: DQO: Add TX path") Signed-off-by: Joshua Washington <[email protected]> Reviewed-by: Praveen Kaligineedi <[email protected]> Reviewed-by: Harshitha Ramamurthy <[email protected]> Reviewed-by: Willem de Bruijn <[email protected]> Suggested-by: Eric Dumazet <[email protected]> Acked-by: Andrei Vagin <[email protected]>
v2 - Remove unnecessary comments, remove line break between fixes tag
and signoffs.
Jakub Kicinski [Wed, 12 Jun 2024 02:40:27 +0000 (19:40 -0700)]
Merge tag 'for-net-2024-06-10' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth
Luiz Augusto von Dentz says:
====================
bluetooth pull request for net:
- hci_sync: fix not using correct handle
- L2CAP: fix rejecting L2CAP_CONN_PARAM_UPDATE_REQ
- L2CAP: fix connection setup in l2cap_connect
* tag 'for-net-2024-06-10' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth:
Bluetooth: fix connection setup in l2cap_connect
Bluetooth: L2CAP: Fix rejecting L2CAP_CONN_PARAM_UPDATE_REQ
Bluetooth: hci_sync: Fix not using correct handle
====================
Linus Torvalds [Tue, 11 Jun 2024 19:04:21 +0000 (12:04 -0700)]
Merge tag 'vfs-6.10-rc4.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs fixes from Christian Brauner:
"Misc:
- Restore debugfs behavior of ignoring unknown mount options
- Fix kernel doc for netfs_wait_for_oustanding_io()
- Fix struct statx comment after new addition for this cycle
- Fix a check in find_next_fd()
iomap:
- Fix data zeroing behavior when an extent spans the block that
contains i_size
- Restore i_size increasing in iomap_write_end() for now to avoid
stale data exposure on xfs with a realtime device
Cachefiles:
- Remove unneeded fdtable.h include
- Improve trace output for cachefiles_obj_{get,put}_ondemand_fd()
- Remove requests from the request list to prevent accessing already
freed requests
- Fix UAF when issuing restore command while the daemon is still
alive by adding an additional reference count to requests
- Fix UAF by grabbing a reference during xarray lookup with xa_lock()
held
- Simplify error handling in cachefiles_ondemand_daemon_read()
- Add consistency checks read and open requests to avoid crashes
- Add a spinlock to protect ondemand_id variable which is used to
determine whether an anonymous cachefiles fd has already been
closed
- Make on-demand reads killable allowing to handle broken cachefiles
daemon better
- Flush all requests after the kernel has been marked dead via
CACHEFILES_DEAD to avoid hung-tasks
- Ensure that closed requests are marked as such to avoid reusing
them with a reopen request
- Defer fd_install() until after copy_to_user() succeeded and thereby
get rid of having to use close_fd()
- Ensure that anonymous cachefiles on-demand fds are reused while
they are valid to avoid pinning already freed cookies"
* tag 'vfs-6.10-rc4.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
iomap: Fix iomap_adjust_read_range for plen calculation
iomap: keep on increasing i_size in iomap_write_end()
cachefiles: remove unneeded include of <linux/fdtable.h>
fs/file: fix the check in find_next_fd()
cachefiles: make on-demand read killable
cachefiles: flush all requests after setting CACHEFILES_DEAD
cachefiles: Set object to close if ondemand_id < 0 in copen
cachefiles: defer exposing anon_fd until after copy_to_user() succeeds
cachefiles: never get a new anonymous fd if ondemand_id is valid
cachefiles: add spin_lock for cachefiles_ondemand_info
cachefiles: add consistency check for copen/cread
cachefiles: remove err_put_fd label in cachefiles_ondemand_daemon_read()
cachefiles: fix slab-use-after-free in cachefiles_ondemand_daemon_read()
cachefiles: fix slab-use-after-free in cachefiles_ondemand_get_fd()
cachefiles: remove requests from xarray during flushing requests
cachefiles: add output string to cachefiles_obj_[get|put]_ondemand_fd
statx: Update offset commentary for struct statx
netfs: fix kernel doc for nets_wait_for_outstanding_io()
debugfs: continue to ignore unknown mount options
netfilter: Use flowlabel flow key when re-routing mangled packets
'ip6 dscp set $v' in an nftables outpute route chain has no effect.
While nftables does detect the dscp change and calls the reroute hook.
But ip6_route_me_harder never sets the dscp/flowlabel:
flowlabel/dsfield routing rules are ignored and no reroute takes place.
Thanks to Yi Chen for an excellent reproducer script that I used
to validate this change.
netfilter: ipset: Fix race between namespace cleanup and gc in the list:set type
Lion Ackermann reported that there is a race condition between namespace cleanup
in ipset and the garbage collection of the list:set type. The namespace
cleanup can destroy the list:set type of sets while the gc of the set type is
waiting to run in rcu cleanup. The latter uses data from the destroyed set which
thus leads use after free. The patch contains the following parts:
- When destroying all sets, first remove the garbage collectors, then wait
if needed and then destroy the sets.
- Fix the badly ordered "wait then remove gc" for the destroy a single set
case.
- Fix the missing rcu locking in the list:set type in the userspace test
case.
- Use proper RCU list handlings in the list:set type.
The patch depends on c1193d9bbbd3 (netfilter: ipset: Add list flush to cancel_gc).
Davide Ornaghi [Wed, 5 Jun 2024 11:03:45 +0000 (13:03 +0200)]
netfilter: nft_inner: validate mandatory meta and payload
Check for mandatory netlink attributes in payload and meta expression
when used embedded from the inner expression, otherwise NULL pointer
dereference is possible from userspace.
Fixes: a150d122b6bd ("netfilter: nft_meta: add inner match support") Fixes: 3a07327d10a0 ("netfilter: nft_inner: support for inner tunnel header matching") Signed-off-by: Davide Ornaghi <[email protected]> Signed-off-by: Pablo Neira Ayuso <[email protected]>
Eric Dumazet [Fri, 7 Jun 2024 12:56:52 +0000 (12:56 +0000)]
tcp: use signed arithmetic in tcp_rtx_probe0_timed_out()
Due to timer wheel implementation, a timer will usually fire
after its schedule.
For instance, for HZ=1000, a timeout between 512ms and 4s
has a granularity of 64ms.
For this range of values, the extra delay could be up to 63ms.
For TCP, this means that tp->rcv_tstamp may be after
inet_csk(sk)->icsk_timeout whenever the timer interrupt
finally triggers, if one packet came during the extra delay.
We need to make sure tcp_rtx_probe0_timed_out() handles this case.
Jakub Kicinski [Tue, 11 Jun 2024 02:49:13 +0000 (19:49 -0700)]
Merge branch 'mptcp-various-fixes'
Matthieu Baerts says:
====================
mptcp: various fixes
The different patches here are some unrelated fixes for MPTCP:
- Patch 1 ensures 'snd_una' is initialised on connect in case of MPTCP
fallback to TCP followed by retransmissions before the processing of
any other incoming packets. A fix for v5.9+.
- Patch 2 makes sure the RmAddr MIB counter is incremented, and only
once per ID, upon the reception of a RM_ADDR. A fix for v5.10+.
- Patch 3 doesn't update 'add addr' related counters if the connect()
was not possible. A fix for v5.7+.
- Patch 4 updates the mailmap file to add Geliang's new email address.
====================
Geliang Tang [Fri, 7 Jun 2024 15:01:51 +0000 (17:01 +0200)]
mailmap: map Geliang's new email address
Just like my other email addresses, map my new one to kernel.org
account too.
My new email address uses "last name, first name" format, which is
different from my other email addresses. This mailmap is also used
to indicate that it is actually the same person.
YonglongLi [Fri, 7 Jun 2024 15:01:50 +0000 (17:01 +0200)]
mptcp: pm: update add_addr counters after connect
The creation of new subflows can fail for different reasons. If no
subflow have been created using the received ADD_ADDR, the related
counters should not be updated, otherwise they will never be decremented
for events related to this ID later on.
For the moment, the number of accepted ADD_ADDR is only decremented upon
the reception of a related RM_ADDR, and only if the remote address ID is
currently being used by at least one subflow. In other words, if no
subflow can be created with the received address, the counter will not
be decremented. In this case, it is then important not to increment
pm.add_addr_accepted counter, and not to modify pm.accept_addr bit.
Note that this patch does not modify the behaviour in case of failures
later on, e.g. if the MP Join is dropped or rejected.
The "remove invalid addresses" MP Join subtest has been modified to
validate this case. The broadcast IP address is added before the "valid"
address that will be used to successfully create a subflow, and the
limit is decreased by one: without this patch, it was not possible to
create the last subflow, because:
- the broadcast address would have been accepted even if it was not
usable: the creation of a subflow to this address results in an error,
- the limit of 2 accepted ADD_ADDR would have then been reached.
YonglongLi [Fri, 7 Jun 2024 15:01:49 +0000 (17:01 +0200)]
mptcp: pm: inc RmAddr MIB counter once per RM_ADDR ID
The RmAddr MIB counter is supposed to be incremented once when a valid
RM_ADDR has been received. Before this patch, it could have been
incremented as many times as the number of subflows connected to the
linked address ID, so it could have been 0, 1 or more than 1.
The "RmSubflow" is incremented after a local operation. In this case,
it is normal to tied it with the number of subflows that have been
actually removed.
The "remove invalid addresses" MP Join subtest has been modified to
validate this case. A broadcast IP address is now used instead: the
client will not be able to create a subflow to this address. The
consequence is that when receiving the RM_ADDR with the ID attached to
this broadcast IP address, no subflow linked to this ID will be found.
Paolo Abeni [Fri, 7 Jun 2024 15:01:48 +0000 (17:01 +0200)]
mptcp: ensure snd_una is properly initialized on connect
This is strictly related to commit fb7a0d334894 ("mptcp: ensure snd_nxt
is properly initialized on connect"). It turns out that syzkaller can
trigger the retransmit after fallback and before processing any other
incoming packet - so that snd_una is still left uninitialized.
Address the issue explicitly initializing snd_una together with snd_nxt
and write_seq.
Johannes Berg [Fri, 7 Jun 2024 15:53:32 +0000 (17:53 +0200)]
net/sched: initialize noop_qdisc owner
When the noop_qdisc owner isn't initialized, then it will be 0,
so packets will erroneously be regarded as having been subject
to recursion as long as only CPU 0 queues them. For non-SMP,
that's all packets, of course. This causes a change in what's
reported to userspace, normally noop_qdisc would drop packets
silently, but with this change the syscall returns -ENOBUFS if
RECVERR is also set on the socket.
Fix this by initializing the owner field to -1, just like it
would be for dynamically allocated qdiscs by qdisc_alloc().
Kent Overstreet [Sat, 8 Jun 2024 21:36:24 +0000 (17:36 -0400)]
bcachefs: Add missing synchronize_srcu_expedited() call when shutting down
We use the polling interface to srcu for tracking pending frees; when
shutting down we don't need to wait for an srcu barrier to free them,
but SRCU still gets confused if we shutdown with an outstanding grace
period.
Kent Overstreet [Fri, 7 Jun 2024 00:01:34 +0000 (20:01 -0400)]
bcachefs: Fix reporting of freed objects from key cache shrinker
We count objects as freed when we move them to the srcu-pending lists
because we're doing the equivalent of a kfree_srcu(); the only
difference is managing the pending list ourself means we can allocate
from the pending list.
Kent Overstreet [Wed, 5 Jun 2024 15:08:20 +0000 (11:08 -0400)]
bcachefs: Split out btree_write_submit_wq
Split the workqueues for btree read completions and btree write
submissions; we don't want concurrency control on btree read
completions, but we do want concurrency control on write submissions,
else blocking in submit_bio() will cause a ton of kworkers to be
allocated.
Pauli Virtanen [Sun, 9 Jun 2024 15:06:20 +0000 (18:06 +0300)]
Bluetooth: fix connection setup in l2cap_connect
The amp_id argument of l2cap_connect() was removed in
commit 84a4bb6548a2 ("Bluetooth: HCI: Remove HCI_AMP support")
It was always called with amp_id == 0, i.e. AMP_ID_BREDR == 0x00 (ie.
non-AMP controller). In the above commit, the code path for amp_id != 0
was preserved, although it should have used the amp_id == 0 one.
Restore the previous behavior of the non-AMP code path, to fix problems
with L2CAP connections.
Fixes: 84a4bb6548a2 ("Bluetooth: HCI: Remove HCI_AMP support") Signed-off-by: Pauli Virtanen <[email protected]> Signed-off-by: Luiz Augusto von Dentz <[email protected]>
This removes the bogus check for max > hcon->le_conn_max_interval since
the later is just the initial maximum conn interval not the maximum the
stack could support which is really 3200=4000ms.
In order to pass GAP/CONN/CPUP/BV-05-C one shall probably enter values
of the following fields in IXIT that would cause hci_check_conn_params
to fail:
David S. Miller [Mon, 10 Jun 2024 12:18:09 +0000 (13:18 +0100)]
Merge branch 'geneve-fixes'
Tariq Toukan says:
====================
geneve fixes
This small patchset by Gal provides bug fixes to the geneve tunnels flows.
Patch 1 fixes an incorrect value returned by the inner network header
offset helper.
Patch 2 fixes an issue inside the mlx5e tunneling flow. It 'happened' to
be harmless so far, before applying patch 1.
Series generated against:
commit d30d0e49da71 ("Merge tag 'net-6.10-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net")
====================
Gal Pressman [Thu, 6 Jun 2024 20:32:49 +0000 (23:32 +0300)]
net/mlx5e: Fix features validation check for tunneled UDP (non-VXLAN) packets
Move the vxlan_features_check() call to after we verified the packet is
a tunneled VXLAN packet.
Without this, tunneled UDP non-VXLAN packets (for ex. GENENVE) might
wrongly not get offloaded.
In some cases, it worked by chance as GENEVE header is the same size as
VXLAN, but it is obviously incorrect.
Gal Pressman [Thu, 6 Jun 2024 20:32:48 +0000 (23:32 +0300)]
geneve: Fix incorrect inner network header offset when innerprotoinherit is set
When innerprotoinherit is set, the tunneled packets do not have an inner
Ethernet header.
Change 'maclen' to not always assume the header length is ETH_HLEN, as
there might not be a MAC header.
This resolves issues with drivers (e.g. mlx5, in
mlx5e_tx_tunnel_accel()) who rely on the skb inner network header offset
to be correct, and use it for TX offloads.
Andy Shevchenko [Thu, 6 Jun 2024 16:13:03 +0000 (19:13 +0300)]
net dsa: qca8k: fix usages of device_get_named_child_node()
The documentation for device_get_named_child_node() mentions this
important point:
"
The caller is responsible for calling fwnode_handle_put() on the
returned fwnode pointer.
"
Add fwnode_handle_put() to avoid leaked references.
Fixes: 1e264f9d2918 ("net: dsa: qca8k: add LEDs basic support") Reviewed-by: Simon Horman <[email protected]> Signed-off-by: Andy Shevchenko <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Eric Dumazet [Thu, 6 Jun 2024 15:46:51 +0000 (15:46 +0000)]
tcp: fix race in tcp_v6_syn_recv_sock()
tcp_v6_syn_recv_sock() calls ip6_dst_store() before
inet_sk(newsk)->pinet6 has been set up.
This means ip6_dst_store() writes over the parent (listener)
np->dst_cookie.
This is racy because multiple threads could share the same
parent and their final np->dst_cookie could be wrong.
Move ip6_dst_store() call after inet_sk(newsk)->pinet6
has been changed and after the copy of parent ipv6_pinfo.
Fixes: e994b2f0fb92 ("tcp: do not lock listener to process SYN packets") Signed-off-by: Eric Dumazet <[email protected]> Reviewed-by: Simon Horman <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Ard Biesheuvel [Tue, 4 Jun 2024 21:32:34 +0000 (22:32 +0100)]
ARM: 9405/1: ftrace: Don't assume stack frames are contiguous in memory
The frame pointer unwinder relies on a standard layout of the stack
frame, consisting of (in downward order)
Calling frame:
PC <---------+
LR |
SP |
FP |
.. locals .. |
Callee frame: |
PC |
LR |
SP |
FP ----------+
where after storing its previous value on the stack, FP is made to point
at the location of PC in the callee stack frame, using the canonical
prologue:
The ftrace code assumes that this activation record is pushed first, and
that any stack space for locals is allocated below this. Strict
adherence to this would imply that the caller's value of SP at the time
of the function call can always be obtained by adding 4 to FP (which
points to PC in the callee frame).
However, recent versions of GCC appear to deviate from this rule, and so
the only reliable way to obtain the caller's value of SP is to read it
from the activation record. Since this involves a read from memory
rather than simple arithmetic, we need to use the uaccess API here which
protects against inadvertent data aborts resulting from attempts to
dereference bogus FP values.
The plain uaccess API is ftrace instrumented itself, so to avoid
unbounded recursion, use the __get_kernel_nofault() primitive directly.
David Wei [Thu, 6 Jun 2024 14:59:08 +0000 (07:59 -0700)]
netdevsim: fix backwards compatibility in nsim_get_iflink()
The default ndo_get_iflink() implementation returns the current ifindex
of the netdev. But the overridden nsim_get_iflink() returns 0 if the
current nsim is not linked, breaking backwards compatibility for
userspace that depend on this behaviour.
Fix the problem by returning the current ifindex if not linked to a
peer.
Linus Torvalds [Sun, 9 Jun 2024 16:04:51 +0000 (09:04 -0700)]
Merge tag 'perf-tools-fixes-for-v6.10-2-2024-06-09' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools
Pull perf tools fixes from Arnaldo Carvalho de Melo:
- Update copies of kernel headers, which resulted in support for the
new 'mseal' syscall, SUBVOL statx return mask bit, RISC-V and PPC
prctls, fcntl's DUPFD_QUERY, POSTED_MSI_NOTIFICATION IRQ vector,
'map_shadow_stack' syscall for x86-32.
- Revert perf.data record memory allocation optimization that ended up
causing a regression, work is being done to re-introduce it in the
next merge window.
- Fix handling of minimal vmlinux.h file used with BPF's CO-RE when
interrupting the build.
* tag 'perf-tools-fixes-for-v6.10-2-2024-06-09' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools:
perf bpf: Fix handling of minimal vmlinux.h file when interrupting the build
Revert "perf record: Reduce memory for recording PERF_RECORD_LOST_SAMPLES event"
tools headers arm64: Sync arm64's cputype.h with the kernel sources
tools headers uapi: Sync linux/stat.h with the kernel sources to pick STATX_SUBVOL
tools headers UAPI: Update i915_drm.h with the kernel sources
tools headers UAPI: Sync kvm headers with the kernel sources
tools arch x86: Sync the msr-index.h copy with the kernel sources
tools headers: Update the syscall tables and unistd.h, mostly to support the new 'mseal' syscall
perf trace beauty: Update the arch/x86/include/asm/irq_vectors.h copy with the kernel sources to pick POSTED_MSI_NOTIFICATION
perf beauty: Update copy of linux/socket.h with the kernel sources
tools headers UAPI: Sync fcntl.h with the kernel sources to pick F_DUPFD_QUERY
tools headers UAPI: Sync linux/prctl.h with the kernel sources
tools include UAPI: Sync linux/stat.h with the kernel sources
Linus Torvalds [Sun, 9 Jun 2024 15:49:13 +0000 (08:49 -0700)]
Merge tag 'edac_urgent_for_v6.10_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras
Pull EDAC fixes from Borislav Petkov:
- Convert PCI core error codes to proper error numbers since latter get
propagated all the way up to the module loading functions
* tag 'edac_urgent_for_v6.10_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
EDAC/igen6: Convert PCIBIOS_* return codes to errnos
EDAC/amd64: Convert PCIBIOS_* return codes to errnos
Commit 070246e4674b ("net: stmmac: Fix for mismatched host/device DMA
address width") added support in the stmmac driver for platform drivers
to indicate the host DMA width, but left it up to authors of the
specific platforms to indicate if their width differed from the addr64
register read from the MAC itself.
Qualcomm's EMAC4 integration supports only up to 36 bit width (as
opposed to the addr64 register indicating 40 bit width). Let's indicate
that in the platform driver to avoid a scenario where the driver will
allocate descriptors of size that is supported by the CPU which in our
case is 36 bit, but as the addr64 register is still capable of 40 bits
the device will use two descriptors as one address.
Fixes: 8c4d92e82d50 ("net: stmmac: dwmac-qcom-ethqos: add support for emac4 on sa8775p platforms") Signed-off-by: Sagar Cheluvegowda <[email protected]> Reviewed-by: Simon Horman <[email protected]> Reviewed-by: Andrew Halaney <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Linus Torvalds [Sun, 9 Jun 2024 02:14:02 +0000 (19:14 -0700)]
Merge tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux
Pull clk fix from Stephen Boyd:
"One fix for the SiFive PRCI clocks so that the device boots again.
This driver was registering clkdev lookups that were always going to
be useless. This wasn't a problem until clkdev started returning an
error in these cases, causing this driver to fail probe, and thus boot
to fail because clks are essential for most drivers. The fix is
simple, don't use clkdev because this is a DT based system where
clkdev isn't used"
* tag 'clk-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux:
clk: sifive: Do not register clkdevs for PRCI clocks
Linus Torvalds [Sun, 9 Jun 2024 02:07:18 +0000 (19:07 -0700)]
Merge tag '6.10-rc2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6
Pull smb client fixes from Steve French:
"Two small smb3 client fixes:
- fix deadlock in umount
- minor cleanup due to netfs change"
* tag '6.10-rc2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: Don't advance the I/O iterator before terminating subrequest
smb: client: fix deadlock in smb2_find_smb_tcon()
projects / linux.git / log