From: David Daney on
On 08/03/2010 01:54 PM, Maciej W. Rozycki wrote:
> On Tue, 3 Aug 2010, David Daney wrote:
>
>> diff --git a/arch/mips/include/asm/inst.h b/arch/mips/include/asm/inst.h
>> index 6489f00..444ff71 100644
>> --- a/arch/mips/include/asm/inst.h
>> +++ b/arch/mips/include/asm/inst.h
>> @@ -247,6 +247,12 @@ struct ma_format { /* FPU multipy and add format (MIPS IV) */
>> unsigned int fmt : 2;
>> };
>>
>> +struct b_format { /* BREAK and SYSCALL */
>> + unsigned int opcode:6;
>> + unsigned int code:20;
>> + unsigned int func:6;
>> +};
>> +
>> #elif defined(__MIPSEL__)
>>
>> struct j_format { /* Jump format */
>
> Please note the code field of the BREAK instruction is by toolchain
> convention (bug-compatibility with the original MIPS assembler or
> suchlike) treated as a pair of swapped 10-bit fields -- you may want to
> double-check consistency of interpretation with usage elsewhere.
>

Indeed, I am familiar with that fact. From patch 3/5 we have:

..
..
..
+void __kprobes jprobe_return(void)
+{
+ /* Assembler quirk necessitates this '0,code' business. */
+ asm volatile(
+ "break 0,%0\n\t"
+ ".globl jprobe_return_end\n"
+ "jprobe_return_end:\n"
+ : : "n" (BRK_KPROBE_BP) : "memory");
+}
..
..
..

The 'break 0,code' construct causes gas to emit values that are
compatible with the other use of struct b_format in the patch set.

David Daney
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Ralf Baechle on
Thanks, applied.

Ralf
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/