[PATCH -fixes] riscv: Fix RISCV_ALTERNATIVE_EARLY

[Alexandre Ghiti]

Hi Conor,

On 27/08/2024 10:38, Conor Dooley wrote:
> On Mon, Aug 26, 2024 at 12:57:37PM +0200, Alexandre Ghiti wrote:
>> RISCV_ALTERNATIVE_EARLY will issue sbi_ecall() very early in the boot
>> process, before the first memory mapping is setup so we can't have any
>> instrumentation happening here.
>>
>> In addition, when the kernel is relocatable, we must also not issue any
>> relocation this early since they would have been patched virtually only.
>>
>> So, instead of disabling instrumentation for the whole kernel/sbi.c file
>> and compiling it with -fno-pie, simply move __sbi_ecall() and
>> __sbi_base_ecall() into their own file where this is fixed.
> IOW, this should fix the issue that we discussed here
> https://lore.kernel.org/linux-riscv/abec162e-f3f2-488c-83d9-be17257a5df8@ghiti.fr/
> also?


Yes, as a side effect it also fixes TRACEPOINTS + [KASAN|RELOCATABLE] as 
I reported in this thread since I moved the tracepoints in a file where 
instrumentation is disabled and compiled with no-pie.


> I'm sorry I didn't get to test that yet, we had some pretty bad IT
> issues in the office the last weeks and I have been avoiding going
> there. I'll try to test this one instead..
>
> _______________________________________________
> linux-riscv mailing list
> linux-riscv at lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-riscv