linux-toradex.git/arch/riscv/kernel/vdso, branch master

riscv: vdso: Add symbols for the alternative section boundaries

2026-08-08T01:24:33+00:00

Currently the alternative sections are extracted from the vDSO
binaries at runtime. This has runtime overhead and also doesn't
work for the compat vDSO.

Extract the offsets of the alternative section during the build with the
existing symbol extraction machinery.

Also add dummy symbols as fallback.

Signed-off-by: Thomas Weißschuh 
Reviewed-by: Nam Cao 
Link: https://patch.msgid.link/20260630-riscv-vdso32-alternative-v1-1-a32fd89b7b1c@linutronix.de
Signed-off-by: Paul Walmsley

Revert "riscv: hwprobe: Fix stale vDSO data for late-initialized keys at boot"

2026-08-08T01:24:30+00:00

This reverts commit 5d15d2ad36b0 ("riscv: hwprobe: Fix stale vDSO data for
late-initialized keys at boot"). The commit ensures synchronization between
the unaligned vector access speed probe kthread and vDSO data read. But now
that the kthread has been removed, this commit can be reverted.

Signed-off-by: Nam Cao 
Tested-by: Anirudh Srinivasan 
Link: https://patch.msgid.link/50ca78a649faf53f8941bc94c9cf8268b3644d38.1781666867.git.namcao@linutronix.de
Signed-off-by: Paul Walmsley

riscv: vdso: Simplify cflags remove logic

2026-08-08T01:24:30+00:00

Listing all flags for each object file is tedious and error-prone.

Replace it with a simpler solution.

Link: https://lore.kernel.org/all/20260630135316-f26f0e0f-c08c-4d4d-9963-10f9985a7689@linutronix.de/
Signed-off-by: Thomas Weißschuh 
Link: https://patch.msgid.link/20260701-riscv-vdso-lto-v1-2-89db0cd82077@linutronix.de
Signed-off-by: Paul Walmsley

arch/riscv: vdso: remove CFI landing pad from rt_sigreturn

2026-07-15T01:34:38+00:00

When CONFIG_RISCV_USER_CFI is enabled, the CFI version of the vDSO, has
a CFI landing pad instruction at the start of __vdso_rt_sigreturn. This
breaks libgcc's unwinding code which matches on the first two
instructions. Other unwinders that rely on similar instruction matching
may also be affected.

Since __vdso_rt_sigreturn is reached as part of signal-return handling
rather than via an indirect call/jump from userspace, it does not need a
CFI landing pad. Remove it and restore the instruction sequence expected
by existing unwinding code.

This matches what was done on arm64 in commit 9a964285572b ("arm64:
vdso: Don't prefix sigreturn trampoline with a BTI C instruction") for a
similar issue.

Cc: stable@vger.kernel.org
Fixes: 37f57bd3faea ("arch/riscv: compile vdso with landing pad and shadow stack note")
Co-authored-by: Joel Stanley 
Signed-off-by: Aurelien Jarno 
Signed-off-by: Joel Stanley 
Link: https://patch.msgid.link/20260623204058.498120-1-aurelien@aurel32.net
[pjw@kernel.org: fixed comment style]
Signed-off-by: Paul Walmsley

riscv: vdso: Do not use LTO for the vDSO

2026-07-13T18:08:27+00:00

With LTO enabled the compiler assumes that the vDSO functions are not
used and optimizes them away completely. Currently this happens to
__vdso_clock_getres(), __vdso_clock_gettime(), __vdso_getrandom(),
__vdso_gettimeofday() and __vdso_riscv_hwprobe().

Disable LTO for the vDSO, as these functions are hand-optimized anyways.

Reported-by: kernel test robot 
Closes: https://lore.kernel.org/oe-kbuild-all/202606301855.WvkSC4kD-lkp@intel.com/
Fixes: 021d23428bdb ("RISC-V: build: Allow LTO to be selected")
Cc: stable@vger.kernel.org
Signed-off-by: Thomas Weißschuh 
Link: https://patch.msgid.link/20260701-riscv-vdso-lto-v1-1-89db0cd82077@linutronix.de
Signed-off-by: Paul Walmsley

riscv: Add build salt to the vDSO

2026-06-25T18:05:05+00:00

The vDSO needs to have a unique build id in a similar manner
to the kernel and modules. Use the build salt macro.

Signed-off-by: Bastian Blank 
Reviewed-by: Nam Cao 
Link: https://patch.msgid.link/ajQY7n0an0YwQ--j@steamhammer.waldi.eu.org
Signed-off-by: Paul Walmsley

arch/riscv: add dual vdso creation logic and select vdso based on hw

2026-01-29T09:38:40+00:00

Shadow stack instructions are taken from the Zimop ISA extension,
which is mandated on RVA23. Any userspace with shadow stack
instructions in it will fault on hardware that doesn't have support
for Zimop.  Thus, a shadow stack-enabled userspace can't be run on
hardware that doesn't support Zimop.

It's not known how Linux userspace providers will respond to this kind
of binary fragmentation.  In order to keep kernel portable across
different hardware, 'arch/riscv/kernel/vdso_cfi' is created which has
Makefile logic to compile 'arch/riscv/kernel/vdso' sources with CFI
flags, and 'arch/riscv/kernel/vdso.c' is modified to select the
appropriate vdso depending on whether the underlying CPU implements
the Zimop extension. Since the offset of vdso symbols will change due
to having two different vdso binaries, there is added logic to include
a new generated vdso offset header and dynamically select the offset
(like for rt_sigreturn).

Signed-off-by: Deepak Gupta 
Acked-by: Charles Mirabile 
Tested-by: Andreas Korb  # QEMU, custom CVA6
Tested-by: Valentin Haudiquet 
Link: https://patch.msgid.link/20251112-v5_user_cfi_series-v23-24-b55691eacf4f@rivosinc.com
[pjw@kernel.org: cleaned up patch description]
Signed-off-by: Paul Walmsley

arch/riscv: compile vdso with landing pad and shadow stack note

2026-01-29T09:38:40+00:00

User mode tasks compiled with Zicfilp may call indirectly into the
vdso (like hwprobe indirect calls). Add support for compiling landing
pads into the vdso. Landing pad instructions in the vdso will be
no-ops for tasks which have not enabled landing pads. Furthermore, add
support for the C sources of the vdso to be compiled with shadow stack
and landing pads enabled as well.

Landing pad and shadow stack instructions are emitted only when the
VDSO_CFI cflags option is defined during compile.

Signed-off-by: Jim Shu 
Reviewed-by: Zong Li 
Signed-off-by: Deepak Gupta 
Tested-by: Andreas Korb  # QEMU, custom CVA6
Tested-by: Valentin Haudiquet 
Link: https://patch.msgid.link/20251112-v5_user_cfi_series-v23-23-b55691eacf4f@rivosinc.com
[pjw@kernel.org: cleaned up patch description, issues reported by checkpatch]
Signed-off-by: Paul Walmsley

riscv: hwprobe: Fix stale vDSO data for late-initialized keys at boot

2025-10-18T04:23:11+00:00

The hwprobe vDSO data for some keys, like MISALIGNED_VECTOR_PERF,
is determined by an asynchronous kthread. This can create a race
condition where the kthread finishes after the vDSO data has
already been populated, causing userspace to read stale values.

To fix this race, a new 'ready' flag is added to the vDSO data,
initialized to 'false' during arch_initcall_sync. This flag is
checked by both the vDSO's user-space code and the riscv_hwprobe
syscall. The syscall serves as a one-time gate, using a completion
to wait for any pending probes before populating the data and
setting the flag to 'true', thus ensuring userspace reads fresh
values on its first request.

Reported-by: Tsukasa OI 
Closes: https://lore.kernel.org/linux-riscv/760d637b-b13b-4518-b6bf-883d55d44e7f@irq.a4lg.com/
Fixes: e7c9d66e313b ("RISC-V: Report vector unaligned access speed hwprobe")
Cc: Palmer Dabbelt 
Cc: Alexandre Ghiti 
Cc: Olof Johansson 
Cc: stable@vger.kernel.org
Reviewed-by: Alexandre Ghiti 
Co-developed-by: Palmer Dabbelt 
Signed-off-by: Palmer Dabbelt 
Signed-off-by: Jingwei Wang 
Link: https://lore.kernel.org/r/20250811142035.105820-1-wangjingwei@iscas.ac.cn
[pjw@kernel.org: fix checkpatch issues]
Signed-off-by: Paul Walmsley

riscv: vdso: Exclude .rodata from the PT_DYNAMIC segment

2025-06-10T20:06:48+00:00

.rodata is implicitly included in the PT_DYNAMIC segment due to
inheriting the segment of the preceding .dynamic section (in both GNU ld
and LLD).  When the .rodata section's size is not a multiple of 16
bytes on riscv64, llvm-readelf will report a "PT_DYNAMIC dynamic table
is invalid" warning.  Note: in the presence of the .dynamic section, GNU
readelf and llvm-readelf's -d option decodes the dynamic section using
the section.

This issue arose after commit 8f8c1ff879fab60f80f3a7aec3000f47e5b03ba9
("riscv: vdso.lds.S: remove hardcoded 0x800 .text start addr"), which
placed .rodata directly after .dynamic by removing .eh_frame.

This patch resolves the implicit inclusion into PT_DYNAMIC by explicitly
specifying the :text output section phdr.

Reported-by: Nathan Chancellor 
Closes: https://github.com/ClangBuiltLinux/linux/issues/2093
Signed-off-by: Fangrui Song 
Tested-by: Nathan Chancellor 
Link: https://lore.kernel.org/r/20250602-riscv-vdso-v1-1-0620cf63cff0@maskray.me
Signed-off-by: Palmer Dabbelt