diff options
author | Will Deacon <will@kernel.org> | 2022-12-15 17:59:12 +0000 |
---|---|---|
committer | Will Deacon <will@kernel.org> | 2022-12-15 17:59:12 +0000 |
commit | c0cd1d541704c45030cbb2031612fdd68e8e15d5 (patch) | |
tree | af47652c54d0a26de2409fed20ab76f705e0340e /arch/arm64/Kconfig | |
parent | 38792972de4294163f44d6360fd221e6f2c22a05 (diff) | |
download | lwn-c0cd1d541704c45030cbb2031612fdd68e8e15d5.tar.gz lwn-c0cd1d541704c45030cbb2031612fdd68e8e15d5.zip |
Revert "arm64: errata: Workaround possible Cortex-A715 [ESR|FAR]_ELx corruption"
This reverts commit 44ecda71fd8a70185c270f5914ac563827fe1d4c.
All versions of this patch on the mailing list, including the version
that ended up getting merged, have portions of code guarded by the
non-existent CONFIG_ARM64_WORKAROUND_2645198 option. Although Anshuman
says he tested the code with some additional debug changes [1], I'm
hesitant to fix the CONFIG option and light up a bunch of code right
before I (and others) disappear for the end of year holidays, during
which time we won't be around to deal with any fallout.
So revert the change for now. We can bring back a fixed, tested version
for a later -rc when folks are thinking about things other than trees
and turkeys.
[1] https://lore.kernel.org/r/b6f61241-e436-5db1-1053-3b441080b8d6@arm.com
Reported-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Link: https://lore.kernel.org/r/20221215094811.23188-1-lukas.bulwahn@gmail.com
Signed-off-by: Will Deacon <will@kernel.org>
Diffstat (limited to 'arch/arm64/Kconfig')
-rw-r--r-- | arch/arm64/Kconfig | 16 |
1 files changed, 0 insertions, 16 deletions
diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c6d15dd6cae3..1ccf7e0653a3 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -971,22 +971,6 @@ config ARM64_ERRATUM_2457168 If unsure, say Y. -config ARM64_ERRATUM_2645198 - bool "Cortex-A715: 2645198: Workaround possible [ESR|FAR]_ELx corruption" - default y - help - This option adds the workaround for ARM Cortex-A715 erratum 2645198. - - If a Cortex-A715 cpu sees a page mapping permissions change from executable - to non-executable, it may corrupt the ESR_ELx and FAR_ELx registers on the - next instruction abort caused by permission fault. - - Only user-space does executable to non-executable permission transition via - mprotect() system call. Workaround the problem by doing a break-before-make - TLB invalidation, for all changes to executable user space mappings. - - If unsure, say Y. - config CAVIUM_ERRATUM_22375 bool "Cavium erratum 22375, 24313" default y |