Skip to contentSkip to conversation

powerpc/64s: Mask SRR0 before checking against the masked NIP

PATCH Applied Nicholas Piggin 2 participants 2 messages Jan 17, 2022 – Jan 19, 2022 PowerPC

lore
Share & tools
lore

Nicholas Piggin

author reviewer
#0 lore raw

PATCH 1/1 powerpc/64s: Mask SRR0 before checking against the masked NIP

Commit 314f6c23dd8d ("powerpc/64s: Mask NIP before checking against SRR0") masked off the low 2 bits of the NIP value in the interrupt stack frame in case they are non-zero and mis-compare against a SRR0 register value of a CPU which always reads back 0 from the 2 low bits which are reserved.

This now causes the opposite problem that an implementation which does implement those bits in SRR0 will mis-compare against the masked NIP value in which they have been cleared. QEMU is one such implementation, and this is allowed by the architecture.

This can be triggered by sigfuz by setting low bits of PT_NIP in the signal context.

Fix this for now by masking the SRR0 bits as well. Cleaner is probably to sanitise these values before putting them in registers or stack, but this is the quick and backportable fix.

Fixes 314f6c23dd8d ("powerpc/64s: Mask NIP before checking against SRR0")
---
 arch/powerpc/kernel/interrupt_64.S | 2 ++
 1 file changed, 2 insertions(+)
arch/powerpc/kernel/interrupt_64.S +2 -0
index 92088f848266..7bab2d7de372 100644
@@ -30,6 +30,7 @@ COMPAT_SYS_CALL_TABLE:
 	.ifc \srr,srr
 	mfspr	r11,SPRN_SRR0
 	ld	r12,_NIP(r1)
+	clrrdi  r11,r11,2
 	clrrdi  r12,r12,2
 100:	tdne	r11,r12
 	EMIT_WARN_ENTRY 100b,__FILE__,__LINE__,(BUGFLAG_WARNING | BUGFLAG_ONCE)
@@ -40,6 +41,7 @@ COMPAT_SYS_CALL_TABLE:
 	.else
 	mfspr	r11,SPRN_HSRR0
 	ld	r12,_NIP(r1)
+	clrrdi  r11,r11,2
 	clrrdi  r12,r12,2
 100:	tdne	r11,r12
 	EMIT_WARN_ENTRY 100b,__FILE__,__LINE__,(BUGFLAG_WARNING | BUGFLAG_ONCE)
Signature · 1 line
2.23.0
#1 lore raw
Quoted 6 lines · Nicholas Piggin

Commit 314f6c23dd8d ("powerpc/64s: Mask NIP before checking against SRR0") masked off the low 2 bits of the NIP value in the interrupt stack frame in case they are non-zero and mis-compare against a SRR0 register value of a CPU which always reads back 0 from the 2 low bits which are reserved.

Quoted 6 lines

This now causes the opposite problem that an implementation which does implement those bits in SRR0 will mis-compare against the masked NIP value in which they have been cleared. QEMU is one such implementation, and this is allowed by the architecture.

[...]

Applied to powerpc/fixes.

[1/1] powerpc/64s: Mask SRR0 before checking against the masked NIP

      https://git.kernel.org/powerpc/c/aee101d7b95a03078945681dd7f7ea5e4a1e7686

cheers