On Thu, Nov 26, 2020, 5:03 PM Michael Ellerman <mpe@ellerman.id.au> wrote:

Bill Wendling <morbo@google.com> = writes:
> On Mon, Nov 23, 2020 at 7:44 PM Michael Ellerman <mpe@ellerman.id.a= u> wrote:
>> Bill Wendling <morbo@google.com> writes:
>> > On Mon, Nov 23, 2020 at 12:10 PM Segher Boessenkool
>> > <segher@kernel.crashing.org> wrote:
>> >> On Mon, Nov 23, 2020 at 12:01:01PM -0800, Bill Wendling w= rote:
>> >> > On Mon, Nov 23, 2020 at 11:58 AM Segher Boessenkool<= br> >> >> > <segher@kernel.crashing.org> wrote= :
>> >> > > > On Sun, Nov 22, 2020 at 10:36 PM Segher Bo= essenkool
>> >> > > > <segher@kernel.crashing.org= > wrote:
>> >> > > > > "true" (as a result of a co= mparison) in as is -1, not 1.
>> >> > >
>> >> > > On Mon, Nov 23, 2020 at 11:43:11AM -0800, Bill = Wendling wrote:
>> >> > > > What Segher said. :-) Also, if you reverse= the comparison, you'll get
>> >> > > > a build error.
>> >> > >
>> >> > > But that means your patch is the wrong way arou= nd?
>> >> > >
>> >> > > -=C2=A0 =C2=A0 =C2=A0 =C2=A0.ifgt (label##4b- l= abel##3b)-(label##2b- label##1b);=C2=A0 =C2=A0 \
>> >> > > -=C2=A0 =C2=A0 =C2=A0 =C2=A0.error "Featur= e section else case larger than body";=C2=A0 =C2=A0 \
>> >> > > -=C2=A0 =C2=A0 =C2=A0 =C2=A0.endif;=C2=A0 =C2= =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0 =C2=A0\
>> >> > > +=C2=A0 =C2=A0 =C2=A0 =C2=A0.org . - ((label##4= b-label##3b) > (label##2b-label##1b)); \
>> >> > >
>> >> > > It should be a + in that last line, not a -. >> >> >
>> >> > I said so in a follow up email.
>> >>
>> >> Yeah, and that arrived a second after I pressed "sen= d" :-)
>> >>
>> > Michael, I apologize for the churn with these patches. I beli= eve the
>> > policy is to resend the match as "v4", correct?
>> >
>> > I ran tests with the change above. It compiled with no error.= If I
>> > switch the labels around to ".org . + ((label##2b-label#= #1b) >
>> > (label##4b-label##3b))", then it fails as expected.
>>
>> I wanted to retain the nicer error reporting for gcc builds, so I = did it
>> like this:
>>
>> diff --git a/arch/powerpc/include/asm/feature-fixups.h b/arch/powe= rpc/include/asm/feature-fixups.h
>> index b0af97add751..c4ad33074df5 100644
>> --- a/arch/powerpc/include/asm/feature-fixups.h
>> +++ b/arch/powerpc/include/asm/feature-fixups.h
>> @@ -36,6 +36,24 @@ label##2:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 \
>>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0.align 2;=C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0\
>>=C2=A0 label##3:
>>
>> +
>> +#ifndef CONFIG_CC_IS_CLANG
>> +#define CHECK_ALT_SIZE(else_size, body_size)=C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0\
>> +=C2=A0 =C2=A0 =C2=A0 =C2=A0.ifgt (else_size) - (body_size);=C2=A0= =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 \
>> +=C2=A0 =C2=A0 =C2=A0 =C2=A0.error "Feature section else case= larger than body";=C2=A0 =C2=A0 \
>> +=C2=A0 =C2=A0 =C2=A0 =C2=A0.endif;
>> +#else
>> +/*
>> + * If we use the ifgt syntax above, clang's assembler complai= ns about the
>> + * expression being non-absolute when the code appears in an inli= ne assembly
>> + * statement.
>> + * As a workaround use an .org directive that has no effect if th= e else case
>> + * instructions are smaller than the body, but fails otherwise. >> + */
>> +#define CHECK_ALT_SIZE(else_size, body_size)=C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0\
>> +=C2=A0 =C2=A0 =C2=A0 =C2=A0.org . + ((else_size) > (body_size)= );
>> +#endif
>> +
>>=C2=A0 #define MAKE_FTR_SECTION_ENTRY(msk, val, label, sect)=C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 \
>>=C2=A0 label##4:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 \
>>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0.popsection;=C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 \
>> @@ -48,9 +66,7 @@ label##5:=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0\
>>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0FTR_ENTRY_OFFSET label##2b-label#= #5b;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0\<= br> >>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0FTR_ENTRY_OFFSET label##3b-label#= #5b;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0\<= br> >>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0FTR_ENTRY_OFFSET label##4b-label#= #5b;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0\<= br> >> -=C2=A0 =C2=A0 =C2=A0 =C2=A0.ifgt (label##4b- label##3b)-(label##2= b- label##1b);=C2=A0 =C2=A0 \
>> -=C2=A0 =C2=A0 =C2=A0 =C2=A0.error "Feature section else case= larger than body";=C2=A0 =C2=A0 \
>> -=C2=A0 =C2=A0 =C2=A0 =C2=A0.endif;=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 = =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0\
>> +=C2=A0 =C2=A0 =C2=A0 =C2=A0CHECK_ALT_SIZE((label##4b-label##3b), = (label##2b-label##1b)); \
>>=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0.popsection;
>>
>>
>>
>> I've pushed a branch with all your patches applied to:
>>
>>=C2=A0 =C2=A0https://github.co= m/linuxppc/linux/commits/next-test
>>
> This works for me. Thanks!

Great.

>> Are you able to give that a quick test? It builds clean with clang= for
>> me, but we must be using different versions of clang because my br= anch
>> already builds clean for me even without your patches.
>>
> You may need to set LLVM_IAS=3D1 to get the behavior I'm seeing. T= hat
> turns on clang's integrated assembler, which I think is disabled b= y
> default.

Yep that does it.

But then I get:
=C2=A0 clang: error: unsupported argument '-mpower4' to option '= ;Wa,'
=C2=A0 clang: error: unsupported argument '-many' to option 'Wa= ,'

So I guess I'm still missing something?

I've seen that too. I'm = not entirely sure what's causing it, but I'll look into it. I'v= e got a backlog of things to work on still. :-) It's probably a clang i= ssue. There's another one that came up having to do with the format of = some PPC instructions. I have a clang fix on review for those.

> Note that with clang's integrated assembler, arch/powerpc/boot/uti= l.S
> fails to compile. Alan Modra mentioned that he sent you a patch to
> "modernize" the file so that clang can compile it.

Ah you're right he did, it didn't go to patchwork so I missed it. H= ave
grabbed it now.

Thanks!