From: Thomas Graziadei <[email protected]>
The 32bit powerpc assembler implementation of the lazy preemption
set the _TIF_PERSYSCALL_MASK on the low word. This could lead to
modprobe segfaults and a kernel panic - not syncing: Attempt to
kill init! issue.
Fixed by shifting the mask by 16 bit using andis and lis.
Signed-off-by: Thomas Graziadei <[email protected]>
---
arch/powerpc/kernel/entry_32.S | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S
index 172dfb567c25..ab609d63d644 100644
--- a/arch/powerpc/kernel/entry_32.S
+++ b/arch/powerpc/kernel/entry_32.S
@@ -533,12 +533,12 @@ syscall_exit_work:
1: stw r6,RESULT(r1) /* Save result */
stw r3,GPR3(r1) /* Update return value */
-2: andi. r0,r9,(_TIF_PERSYSCALL_MASK)@h
+2: andis. r0,r9,(_TIF_PERSYSCALL_MASK)@h
beq 4f
/* Clear per-syscall TIF flags if any are set. */
- li r11,_TIF_PERSYSCALL_MASK@h
+ lis r11,(_TIF_PERSYSCALL_MASK)@h
addi r12,r2,TI_FLAGS
3: lwarx r8,0,r12
andc r8,r8,r11
--
2.17.1
On 2020-03-18 21:26:40 [+0100], Thomas Graziadei wrote:
> From: Thomas Graziadei <[email protected]>
>
> The 32bit powerpc assembler implementation of the lazy preemption
> set the _TIF_PERSYSCALL_MASK on the low word. This could lead to
> modprobe segfaults and a kernel panic - not syncing: Attempt to
> kill init! issue.
>
> Fixed by shifting the mask by 16 bit using andis and lis.
bah. Thank you for catching this.
Still e500 based powerpc I assume?
> Signed-off-by: Thomas Graziadei <[email protected]>
Sebastian
> From: Sebastian Andrzej Siewior [mailto:[email protected]]
> On 2020-03-18 21:26:40 [+0100], Thomas Graziadei wrote:
>> From: Thomas Graziadei <[email protected]>
>>
>> The 32bit powerpc assembler implementation of the lazy preemption set
>> the _TIF_PERSYSCALL_MASK on the low word. This could lead to modprobe
>> segfaults and a kernel panic - not syncing: Attempt to kill init!
>> issue.
>>
>> Fixed by shifting the mask by 16 bit using andis and lis.
>
> bah. Thank you for catching this.
> Still e500 based powerpc I assume?
Well thanks for the great work and yes still e500 powerpc based.
>
>> Signed-off-by: Thomas Graziadei <[email protected]>
>
> Sebastian
Thomas