Re: [PATCH] powerpc/mm: mark more tlb functions as __always_inline

From: Michael Ellerman
Date: Tue May 21 2019 - 03:45:02 EST


Masahiro Yamada <yamada.masahiro@xxxxxxxxxxxxx> writes:
> On Tue, May 21, 2019 at 3:54 PM Christophe Leroy
> <christophe.leroy@xxxxxx> wrote:
>> Le 21/05/2019 Ã 08:16, Masahiro Yamada a Ãcrit :
>> > With CONFIG_OPTIMIZE_INLINING enabled, Laura Abbott reported error
>> > with gcc 9.1.1:
>> >
>> > arch/powerpc/mm/book3s64/radix_tlb.c: In function '_tlbiel_pid':
>> > arch/powerpc/mm/book3s64/radix_tlb.c:104:2: warning: asm operand 3 probably doesn't match constraints
>> > 104 | asm volatile(PPC_TLBIEL(%0, %4, %3, %2, %1)
>> > | ^~~
>> > arch/powerpc/mm/book3s64/radix_tlb.c:104:2: error: impossible constraint in 'asm'
>> >
>> > Fixing _tlbiel_pid() is enough to address the warning above, but I
>> > inlined more functions to fix all potential issues.
>> >
>> > To meet the 'i' (immediate) constraint for the asm operands, functions
>> > propagating propagated 'ric' must be always inlined.
>> >
>> > Fixes: 9012d011660e ("compiler: allow all arches to enable CONFIG_OPTIMIZE_INLINING")
>> > Reported-by: Laura Abbott <labbott@xxxxxxxxxx>
>> > Signed-off-by: Masahiro Yamada <yamada.masahiro@xxxxxxxxxxxxx>
>> > ---
>> >
>> > arch/powerpc/mm/book3s64/hash_native.c | 8 +++--
>> > arch/powerpc/mm/book3s64/radix_tlb.c | 44 +++++++++++++++-----------
>> > 2 files changed, 30 insertions(+), 22 deletions(-)
>> >
>> > diff --git a/arch/powerpc/mm/book3s64/hash_native.c b/arch/powerpc/mm/book3s64/hash_native.c
>> > index aaa28fd918fe..bc2c35c0d2b1 100644
>> > --- a/arch/powerpc/mm/book3s64/hash_native.c
>> > +++ b/arch/powerpc/mm/book3s64/hash_native.c
>> > @@ -60,9 +60,11 @@ static inline void tlbiel_hash_set_isa206(unsigned int set, unsigned int is)
>> > * tlbiel instruction for hash, set invalidation
>> > * i.e., r=1 and is=01 or is=10 or is=11
>> > */
>> > -static inline void tlbiel_hash_set_isa300(unsigned int set, unsigned int is,
>> > - unsigned int pid,
>> > - unsigned int ric, unsigned int prs)
>> > +static __always_inline void tlbiel_hash_set_isa300(unsigned int set,
>> > + unsigned int is,
>> > + unsigned int pid,
>> > + unsigned int ric,
>> > + unsigned int prs)
>>
>> Please don't split the line more than it is.
>>
>> powerpc accepts lines up to 90 chars, see arch/powerpc/tools/checkpatch.pl
>
> Ugh, I did not know this. Horrible.
>
> The Linux coding style should be global in the kernel tree.
> No subsystem should adopts its own coding style.

Well that ship sailed long ago.

But we don't have our own coding style, we just don't enforce 80 columns
rigidly, there are cases where a slightly longer line (up to ~90) is
preferable to a split line.

In a case like this with a long attribute and function name I think this
is probably the least worst option:

static __always_inline
void tlbiel_hash_set_isa300(unsigned int set, unsigned int is, unsigned int pid,
unsigned int ric, unsigned int prs)
{
...

cheers