From: "Nicholas Piggin" <npiggin@gmail.com>
To: "Rohan McLure" <rmclure@linux.ibm.com>, <linuxppc-dev@lists.ozlabs.org>
Subject: Re: [PATCH v6 22/25] powerpc/64s: Clear user GPRs in syscall interrupt entry
Date: Fri, 23 Sep 2022 18:02:01 +1000 [thread overview]
Message-ID: <CN3MGF4GO42K.22AHRCY1C91N2@bobo> (raw)
In-Reply-To: <20220921065605.1051927-23-rmclure@linux.ibm.com>
On Wed Sep 21, 2022 at 4:56 PM AEST, Rohan McLure wrote:
> Clear user state in gprs (assign to zero) to reduce the influence of user
> registers on speculation within kernel syscall handlers. Clears occur
> at the very beginning of the sc and scv 0 interrupt handlers, with
> restores occurring following the execution of the syscall handler.
>
> Signed-off-by: Rohan McLure <rmclure@linux.ibm.com>
> ---
> V2: Update summary
> V3: Remove erroneous summary paragraph on syscall_exit_prepare
> V4: Use ZEROIZE instead of NULLIFY. Clear r0 also.
> V5: Move to end of patch series.
> V6: Include clears which were previously in the syscall wrapper patch.
> Move comment on r3-r8 register save to when we alter the calling
> convention for system_call_exception.
The series looks good to here, I just need to find a bit more time to
look at the code and do some tests with the next few patches. I don't
see much problem with them, looks a lot better now with fewer ifdefs
so that's good. Possibly you could share some of those new sanitize
macros in a header file but that's a minor nit.
Coud we have this zeroize also under the same config option as the
next? I figure if we care about speculative security we want both,
and if we don't we need neither.
Thanks,
Nick
> ---
> arch/powerpc/kernel/interrupt_64.S | 17 ++++++++++++++++-
> 1 file changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/arch/powerpc/kernel/interrupt_64.S b/arch/powerpc/kernel/interrupt_64.S
> index a5dd78bdbe6d..40147558e1a6 100644
> --- a/arch/powerpc/kernel/interrupt_64.S
> +++ b/arch/powerpc/kernel/interrupt_64.S
> @@ -106,6 +106,13 @@ END_FTR_SECTION_IFSET(CPU_FTR_HAS_PPR)
> * but this is the best we can do.
> */
>
> + /*
> + * Zero user registers to prevent influencing speculative execution
> + * state of kernel code.
> + */
> + ZEROIZE_GPR(0)
> + ZEROIZE_GPRS(5, 12)
> + ZEROIZE_NVGPRS()
> bl system_call_exception
>
> .Lsyscall_vectored_\name\()_exit:
> @@ -134,6 +141,7 @@ BEGIN_FTR_SECTION
> HMT_MEDIUM_LOW
> END_FTR_SECTION_IFSET(CPU_FTR_HAS_PPR)
>
> + REST_NVGPRS(r1)
> cmpdi r3,0
> bne .Lsyscall_vectored_\name\()_restore_regs
>
> @@ -285,6 +293,13 @@ END_BTB_FLUSH_SECTION
> wrteei 1
> #endif
>
> + /*
> + * Zero user registers to prevent influencing speculative execution
> + * state of kernel code.
> + */
> + ZEROIZE_GPR(0)
> + ZEROIZE_GPRS(5, 12)
> + ZEROIZE_NVGPRS()
> bl system_call_exception
>
> .Lsyscall_exit:
> @@ -325,6 +340,7 @@ BEGIN_FTR_SECTION
> stdcx. r0,0,r1 /* to clear the reservation */
> END_FTR_SECTION_IFCLR(CPU_FTR_STCX_CHECKS_ADDRESS)
>
> + REST_NVGPRS(r1)
> cmpdi r3,0
> bne .Lsyscall_restore_regs
> /* Zero volatile regs that may contain sensitive kernel data */
> @@ -352,7 +368,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_HAS_PPR)
> .Lsyscall_restore_regs:
> ld r3,_CTR(r1)
> ld r4,_XER(r1)
> - REST_NVGPRS(r1)
> mtctr r3
> mtspr SPRN_XER,r4
> REST_GPR(0, r1)
> --
> 2.34.1
next prev parent reply other threads:[~2022-09-23 8:02 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-21 6:55 [PATCH v6 00/25] powerpc: Syscall wrapper and register clearing Rohan McLure
2022-09-21 6:55 ` [PATCH v6 01/25] powerpc: Remove asmlinkage from syscall handler definitions Rohan McLure
2022-09-21 6:55 ` [PATCH v6 02/25] powerpc: Save caller r3 prior to system_call_exception Rohan McLure
2022-09-21 6:55 ` [PATCH v6 03/25] powerpc: Add ZEROIZE_GPRS macros for register clears Rohan McLure
2022-09-21 6:55 ` [PATCH v6 04/25] powerpc/64s: Use {ZEROIZE,SAVE,REST}_GPRS macros in sc, scv 0 handlers Rohan McLure
2022-09-21 6:55 ` [PATCH v6 05/25] powerpc/32: Clarify interrupt restores with REST_GPR macro in entry_32.S Rohan McLure
2022-09-21 6:55 ` [PATCH v6 06/25] powerpc/64e: Clarify register saves and clears with {SAVE,ZEROIZE}_GPRS Rohan McLure
2022-09-21 6:55 ` [PATCH v6 07/25] powerpc/64s: Fix comment on interrupt handler prologue Rohan McLure
2022-09-21 6:55 ` [PATCH v6 08/25] powerpc: Fix fallocate and fadvise64_64 compat parameter combination Rohan McLure
2022-09-21 6:55 ` [PATCH v6 09/25] asm-generic: compat: Support BE for long long args in 32-bit ABIs Rohan McLure
2022-10-31 13:23 ` [PATCH] asm-generic: compat: fix compat_arg_u64 and compat_arg_u64_dual Andreas Schwab
2022-11-01 12:25 ` Michael Ellerman
2022-11-03 8:20 ` Christophe Leroy
2022-11-03 8:48 ` Arnd Bergmann
2022-09-21 6:55 ` [PATCH v6 10/25] powerpc: Use generic fallocate compatibility syscall Rohan McLure
2022-09-21 6:55 ` [PATCH v6 11/25] powerpc/32: Remove powerpc select specialisation Rohan McLure
2022-09-21 6:55 ` [PATCH v6 12/25] powerpc: Remove direct call to personality syscall handler Rohan McLure
2022-09-21 6:55 ` [PATCH v6 13/25] powerpc: Remove direct call to mmap2 syscall handlers Rohan McLure
2022-09-28 12:15 ` Michael Ellerman
2022-09-28 13:00 ` Arnd Bergmann
2022-09-30 13:19 ` Michael Ellerman
2022-09-30 14:09 ` Arnd Bergmann
2022-09-21 6:55 ` [PATCH v6 14/25] powerpc: Provide do_ppc64_personality helper Rohan McLure
2022-09-21 6:55 ` [PATCH v6 15/25] powerpc: Adopt SYSCALL_DEFINE for arch-specific syscall handlers Rohan McLure
2022-09-21 6:55 ` [PATCH v6 16/25] powerpc: Include all arch-specific syscall prototypes Rohan McLure
2022-09-21 6:55 ` [PATCH v6 17/25] powerpc: Enable compile-time check for syscall handlers Rohan McLure
2022-09-21 6:55 ` [PATCH v6 18/25] powerpc: Use common syscall handler type Rohan McLure
2022-09-21 6:55 ` [PATCH v6 19/25] powerpc: Remove high-order word clearing on compat syscall entry Rohan McLure
2022-09-23 7:40 ` Nicholas Piggin
2022-09-28 11:56 ` Michael Ellerman
2022-09-21 6:56 ` [PATCH v6 20/25] powerpc: Change system_call_exception calling convention Rohan McLure
2022-09-23 7:43 ` Nicholas Piggin
2022-09-21 6:56 ` [PATCH v6 21/25] powerpc: Provide syscall wrapper Rohan McLure
2022-09-23 7:50 ` Nicholas Piggin
2022-10-30 15:34 ` Andreas Schwab
2022-10-30 15:50 ` Andreas Schwab
2022-10-30 19:43 ` Arnd Bergmann
2022-10-30 20:05 ` Andreas Schwab
2022-10-31 3:09 ` Michael Ellerman
2022-10-31 14:47 ` [PATCH] powerpc/32: fix syscall wrappers with 64-bit arguments Andreas Schwab
2022-10-31 19:37 ` Arnd Bergmann
2022-11-01 12:25 ` Michael Ellerman
2022-09-21 6:56 ` [PATCH v6 22/25] powerpc/64s: Clear user GPRs in syscall interrupt entry Rohan McLure
2022-09-23 8:02 ` Nicholas Piggin [this message]
2022-10-31 23:22 ` Rohan McLure
2022-09-21 6:56 ` [PATCH v6 23/25] powerpc/64: Add INTERRUPT_SANITIZE_REGISTERS Kconfig Rohan McLure
2022-09-21 6:56 ` [PATCH v6 24/25] powerpc/64s: Clear gprs on interrupt routine entry in Book3S Rohan McLure
2022-09-21 6:56 ` [PATCH v6 25/25] powerpc/64e: Clear gprs on interrupt routine entry on Book3E Rohan McLure
2022-10-04 13:24 ` [PATCH v6 00/25] powerpc: Syscall wrapper and register clearing Michael Ellerman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CN3MGF4GO42K.22AHRCY1C91N2@bobo \
--to=npiggin@gmail.com \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=rmclure@linux.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).