[PATCH] powerpc/machdep: warn when machine_is() used too early
Michael Ellerman
mpe at ellerman.id.au
Mon Feb 13 21:45:04 AEDT 2023
Christophe Leroy <christophe.leroy at csgroup.eu> writes:
> Le 11/02/2023 à 00:56, Nathan Lynch via B4 Submission Endpoint a écrit :
>> From: Nathan Lynch <nathanl at linux.ibm.com>
>>
>> machine_is() can't provide correct results before probe_machine() has
>> run. Warn when it's used too early in boot.
>>
>> Signed-off-by: Nathan Lynch <nathanl at linux.ibm.com>
>> ---
>> Prompted by my attempts to do some pseries-specific setup during
>> rtas_initialize() and being puzzled for a while that it wasn't
>> working.
>> ---
>> arch/powerpc/include/asm/machdep.h | 12 +++++++-----
>> 1 file changed, 7 insertions(+), 5 deletions(-)
>>
>> diff --git a/arch/powerpc/include/asm/machdep.h b/arch/powerpc/include/asm/machdep.h
>> index 378b8d5836a7..8c0a799d18cd 100644
>> --- a/arch/powerpc/include/asm/machdep.h
>> +++ b/arch/powerpc/include/asm/machdep.h
>> @@ -220,11 +220,13 @@ extern struct machdep_calls *machine_id;
>> EXPORT_SYMBOL(mach_##name); \
>> struct machdep_calls mach_##name __machine_desc =
>>
>> -#define machine_is(name) \
>> - ({ \
>> - extern struct machdep_calls mach_##name \
>> - __attribute__((weak)); \
>> - machine_id == &mach_##name; \
>> +#define machine_is(name) \
>> + ({ \
>> + extern struct machdep_calls mach_##name \
>> + __attribute__((weak)); \
>> + WARN(!machine_id, \
>> + "machine_is() called before probe_machine()"); \
>
> Is a WARN() really necessary ? WARN() is less optimised than WARN_ON(),
> especially on PPC64.
>
> This should never ever happen so a WARN_ON(!machine_id) should be
> enough, the developper that hits it is able to go to the given file:line
> and understand what happened.
Yeah I agree, WARN_ON() should be sufficient here, and should generate
slightly better code. We have > 100 uses of machine_is(), so keeping
each small is desirable.
cheers
More information about the Linuxppc-dev
mailing list