diff mbox series

RISC-V: fix barrier() use in <vdso/processor.h>

Message ID 20201117013951.7827-1-rdunlap@infradead.org (mailing list archive)
State New, archived
Headers show
Series RISC-V: fix barrier() use in <vdso/processor.h> | expand

Commit Message

Randy Dunlap Nov. 17, 2020, 1:39 a.m. UTC
riscv's <vdso/processor.h> uses barrier() so it should
#include <asm/barrier.h> to prevent build errors.

Fixes this build error:
  CC [M]  drivers/net/ethernet/emulex/benet/be_main.o
In file included from ./include/vdso/processor.h:10,
                 from ./arch/riscv/include/asm/processor.h:11,
                 from ./include/linux/prefetch.h:15,
                 from drivers/net/ethernet/emulex/benet/be_main.c:14:
./arch/riscv/include/asm/vdso/processor.h: In function 'cpu_relax':
./arch/riscv/include/asm/vdso/processor.h:14:2: error: implicit declaration of function 'barrier' [-Werror=implicit-function-declaration]
   14 |  barrier();

This happens with a total of 5 networking drivers -- they all use
<linux/prefetch.h>.

rv64 allmodconfig now builds cleanly after this patch.

Fixes fallout from:
815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")

Fixes: ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")
Reported-by: Andreas Schwab <schwab@linux-m68k.org>
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Arvind Sankar <nivedita@alum.mit.edu>
Cc: linux-riscv@lists.infradead.org
Cc: clang-built-linux@googlegroups.com
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Nathan Chancellor <natechancellor@gmail.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Albert Ou <aou@eecs.berkeley.edu>
---
 arch/riscv/include/asm/vdso/processor.h |    2 ++
 1 file changed, 2 insertions(+)

Comments

Nick Desaulniers Nov. 17, 2020, 7 p.m. UTC | #1
On Mon, Nov 16, 2020 at 5:40 PM Randy Dunlap <rdunlap@infradead.org> wrote:
>
> riscv's <vdso/processor.h> uses barrier() so it should
> #include <asm/barrier.h> to prevent build errors.
>
> Fixes this build error:
>   CC [M]  drivers/net/ethernet/emulex/benet/be_main.o
> In file included from ./include/vdso/processor.h:10,
>                  from ./arch/riscv/include/asm/processor.h:11,
>                  from ./include/linux/prefetch.h:15,
>                  from drivers/net/ethernet/emulex/benet/be_main.c:14:
> ./arch/riscv/include/asm/vdso/processor.h: In function 'cpu_relax':
> ./arch/riscv/include/asm/vdso/processor.h:14:2: error: implicit declaration of function 'barrier' [-Werror=implicit-function-declaration]
>    14 |  barrier();
>
> This happens with a total of 5 networking drivers -- they all use
> <linux/prefetch.h>.
>
> rv64 allmodconfig now builds cleanly after this patch.
>
> Fixes fallout from:
> 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")
>
> Fixes: ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")

Hi Randy,
Thanks for the patch, it looks good to me.  I only had a question on
the commit message.

Is this also related to:
commit 3347acc6fcd4 ("compiler.h: fix barrier_data() on clang")

I plan to backport 3347acc6fcd4, so it would be good if all of the
follow ups are all tagged accordingly with fixes tags, so that I don't
break anything backporting 3347acc6fcd4 because I forgot or missed any
follow ups that also needed to be backported.

> Reported-by: Andreas Schwab <schwab@linux-m68k.org>
> Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Stephen Rothwell <sfr@canb.auug.org.au>
> Cc: Arvind Sankar <nivedita@alum.mit.edu>
> Cc: linux-riscv@lists.infradead.org
> Cc: clang-built-linux@googlegroups.com
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Nathan Chancellor <natechancellor@gmail.com>
> Cc: Paul Walmsley <paul.walmsley@sifive.com>
> Cc: Palmer Dabbelt <palmer@dabbelt.com>
> Cc: Albert Ou <aou@eecs.berkeley.edu>
> ---
>  arch/riscv/include/asm/vdso/processor.h |    2 ++
>  1 file changed, 2 insertions(+)
>
> --- lnx-510-rc4.orig/arch/riscv/include/asm/vdso/processor.h
> +++ lnx-510-rc4/arch/riscv/include/asm/vdso/processor.h
> @@ -4,6 +4,8 @@
>
>  #ifndef __ASSEMBLY__
>
> +#include <asm/barrier.h>
> +
>  static inline void cpu_relax(void)
>  {
>  #ifdef __riscv_muldiv
Randy Dunlap Nov. 17, 2020, 7:22 p.m. UTC | #2
On 11/17/20 11:00 AM, Nick Desaulniers wrote:
> On Mon, Nov 16, 2020 at 5:40 PM Randy Dunlap <rdunlap@infradead.org> wrote:
>>
>> riscv's <vdso/processor.h> uses barrier() so it should
>> #include <asm/barrier.h> to prevent build errors.
>>
>> Fixes this build error:
>>   CC [M]  drivers/net/ethernet/emulex/benet/be_main.o
>> In file included from ./include/vdso/processor.h:10,
>>                  from ./arch/riscv/include/asm/processor.h:11,
>>                  from ./include/linux/prefetch.h:15,
>>                  from drivers/net/ethernet/emulex/benet/be_main.c:14:
>> ./arch/riscv/include/asm/vdso/processor.h: In function 'cpu_relax':
>> ./arch/riscv/include/asm/vdso/processor.h:14:2: error: implicit declaration of function 'barrier' [-Werror=implicit-function-declaration]
>>    14 |  barrier();
>>
>> This happens with a total of 5 networking drivers -- they all use
>> <linux/prefetch.h>.
>>
>> rv64 allmodconfig now builds cleanly after this patch.
>>
>> Fixes fallout from:
>> 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")
>>
>> Fixes: ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")
> 
> Hi Randy,
> Thanks for the patch, it looks good to me.  I only had a question on
> the commit message.
> 
> Is this also related to:
> commit 3347acc6fcd4 ("compiler.h: fix barrier_data() on clang")

Hi Nick,
Yes, it looks like I tagged the wrong commit in Fixes:
Thanks for noticing that.


> I plan to backport 3347acc6fcd4, so it would be good if all of the
> follow ups are all tagged accordingly with fixes tags, so that I don't
> break anything backporting 3347acc6fcd4 because I forgot or missed any
> follow ups that also needed to be backported.
> 
>> Reported-by: Andreas Schwab <schwab@linux-m68k.org>
>> Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
>> Cc: Andrew Morton <akpm@linux-foundation.org>
>> Cc: Stephen Rothwell <sfr@canb.auug.org.au>
>> Cc: Arvind Sankar <nivedita@alum.mit.edu>
>> Cc: linux-riscv@lists.infradead.org
>> Cc: clang-built-linux@googlegroups.com
>> Cc: Nick Desaulniers <ndesaulniers@google.com>
>> Cc: Nathan Chancellor <natechancellor@gmail.com>
>> Cc: Paul Walmsley <paul.walmsley@sifive.com>
>> Cc: Palmer Dabbelt <palmer@dabbelt.com>
>> Cc: Albert Ou <aou@eecs.berkeley.edu>
>> ---
>>  arch/riscv/include/asm/vdso/processor.h |    2 ++
>>  1 file changed, 2 insertions(+)
>>
>> --- lnx-510-rc4.orig/arch/riscv/include/asm/vdso/processor.h
>> +++ lnx-510-rc4/arch/riscv/include/asm/vdso/processor.h
>> @@ -4,6 +4,8 @@
>>
>>  #ifndef __ASSEMBLY__
>>
>> +#include <asm/barrier.h>
>> +
>>  static inline void cpu_relax(void)
>>  {
>>  #ifdef __riscv_muldiv
> 
> 
>
Randy Dunlap Nov. 17, 2020, 7:27 p.m. UTC | #3
On 11/17/20 11:22 AM, Randy Dunlap wrote:
> On 11/17/20 11:00 AM, Nick Desaulniers wrote:
>> On Mon, Nov 16, 2020 at 5:40 PM Randy Dunlap <rdunlap@infradead.org> wrote:
>>>
>>> riscv's <vdso/processor.h> uses barrier() so it should
>>> #include <asm/barrier.h> to prevent build errors.
>>>
>>> Fixes this build error:
>>>   CC [M]  drivers/net/ethernet/emulex/benet/be_main.o
>>> In file included from ./include/vdso/processor.h:10,
>>>                  from ./arch/riscv/include/asm/processor.h:11,
>>>                  from ./include/linux/prefetch.h:15,
>>>                  from drivers/net/ethernet/emulex/benet/be_main.c:14:
>>> ./arch/riscv/include/asm/vdso/processor.h: In function 'cpu_relax':
>>> ./arch/riscv/include/asm/vdso/processor.h:14:2: error: implicit declaration of function 'barrier' [-Werror=implicit-function-declaration]
>>>    14 |  barrier();
>>>
>>> This happens with a total of 5 networking drivers -- they all use
>>> <linux/prefetch.h>.
>>>
>>> rv64 allmodconfig now builds cleanly after this patch.
>>>
>>> Fixes fallout from:
>>> 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")
>>>
>>> Fixes: ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")
>>
>> Hi Randy,
>> Thanks for the patch, it looks good to me.  I only had a question on
>> the commit message.
>>
>> Is this also related to:
>> commit 3347acc6fcd4 ("compiler.h: fix barrier_data() on clang")
> 
> Hi Nick,
> Yes, it looks like I tagged the wrong commit in Fixes:
> Thanks for noticing that.
> 

I'll send a v2 with corrections.

> 
>> I plan to backport 3347acc6fcd4, so it would be good if all of the
>> follow ups are all tagged accordingly with fixes tags, so that I don't
>> break anything backporting 3347acc6fcd4 because I forgot or missed any
>> follow ups that also needed to be backported.
>>
>>> Reported-by: Andreas Schwab <schwab@linux-m68k.org>
>>> Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
>>> Cc: Andrew Morton <akpm@linux-foundation.org>
>>> Cc: Stephen Rothwell <sfr@canb.auug.org.au>
>>> Cc: Arvind Sankar <nivedita@alum.mit.edu>
>>> Cc: linux-riscv@lists.infradead.org
>>> Cc: clang-built-linux@googlegroups.com
>>> Cc: Nick Desaulniers <ndesaulniers@google.com>
>>> Cc: Nathan Chancellor <natechancellor@gmail.com>
>>> Cc: Paul Walmsley <paul.walmsley@sifive.com>
>>> Cc: Palmer Dabbelt <palmer@dabbelt.com>
>>> Cc: Albert Ou <aou@eecs.berkeley.edu>
>>> ---
>>>  arch/riscv/include/asm/vdso/processor.h |    2 ++
>>>  1 file changed, 2 insertions(+)
>>>
>>> --- lnx-510-rc4.orig/arch/riscv/include/asm/vdso/processor.h
>>> +++ lnx-510-rc4/arch/riscv/include/asm/vdso/processor.h
>>> @@ -4,6 +4,8 @@
>>>
>>>  #ifndef __ASSEMBLY__
>>>
>>> +#include <asm/barrier.h>
>>> +
>>>  static inline void cpu_relax(void)
>>>  {
>>>  #ifdef __riscv_muldiv
Arvind Sankar Nov. 17, 2020, 7:32 p.m. UTC | #4
On Tue, Nov 17, 2020 at 11:22:39AM -0800, Randy Dunlap wrote:
> On 11/17/20 11:00 AM, Nick Desaulniers wrote:
> > On Mon, Nov 16, 2020 at 5:40 PM Randy Dunlap <rdunlap@infradead.org> wrote:
> >>
> >> riscv's <vdso/processor.h> uses barrier() so it should
> >> #include <asm/barrier.h> to prevent build errors.
> >>
> >> Fixes this build error:
> >>   CC [M]  drivers/net/ethernet/emulex/benet/be_main.o
> >> In file included from ./include/vdso/processor.h:10,
> >>                  from ./arch/riscv/include/asm/processor.h:11,
> >>                  from ./include/linux/prefetch.h:15,
> >>                  from drivers/net/ethernet/emulex/benet/be_main.c:14:
> >> ./arch/riscv/include/asm/vdso/processor.h: In function 'cpu_relax':
> >> ./arch/riscv/include/asm/vdso/processor.h:14:2: error: implicit declaration of function 'barrier' [-Werror=implicit-function-declaration]
> >>    14 |  barrier();
> >>
> >> This happens with a total of 5 networking drivers -- they all use
> >> <linux/prefetch.h>.
> >>
> >> rv64 allmodconfig now builds cleanly after this patch.
> >>
> >> Fixes fallout from:
> >> 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")
> >>
> >> Fixes: ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")
> > 
> > Hi Randy,
> > Thanks for the patch, it looks good to me.  I only had a question on
> > the commit message.
> > 
> > Is this also related to:
> > commit 3347acc6fcd4 ("compiler.h: fix barrier_data() on clang")
> 
> Hi Nick,
> Yes, it looks like I tagged the wrong commit in Fixes:
> Thanks for noticing that.
> 

Thanks for the patch!

Acked-by: Arvind Sankar <nivedita@alum.mit.edu>

Generally speaking, should compiler_types.h be considered "internal"?
i.e. everyone should include compiler.h, not compiler_types.h directly?

Eg for another potential case that might bite is that RELOC_HIDE is
defined in compiler_types.h, except for clang, which will only have a
definition if compiler.h is included.
Palmer Dabbelt Nov. 21, 2020, 3:25 a.m. UTC | #5
On Mon, 16 Nov 2020 17:39:51 PST (-0800), rdunlap@infradead.org wrote:
> riscv's <vdso/processor.h> uses barrier() so it should
> #include <asm/barrier.h> to prevent build errors.
>
> Fixes this build error:
>   CC [M]  drivers/net/ethernet/emulex/benet/be_main.o
> In file included from ./include/vdso/processor.h:10,
>                  from ./arch/riscv/include/asm/processor.h:11,
>                  from ./include/linux/prefetch.h:15,
>                  from drivers/net/ethernet/emulex/benet/be_main.c:14:
> ./arch/riscv/include/asm/vdso/processor.h: In function 'cpu_relax':
> ./arch/riscv/include/asm/vdso/processor.h:14:2: error: implicit declaration of function 'barrier' [-Werror=implicit-function-declaration]
>    14 |  barrier();
>
> This happens with a total of 5 networking drivers -- they all use
> <linux/prefetch.h>.
>
> rv64 allmodconfig now builds cleanly after this patch.
>
> Fixes fallout from:
> 815f0ddb346c ("include/linux/compiler*.h: make compiler-*.h mutually exclusive")
>
> Fixes: ad5d1122b82f ("riscv: use vDSO common flow to reduce the latency of the time-related functions")
> Reported-by: Andreas Schwab <schwab@linux-m68k.org>
> Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: Stephen Rothwell <sfr@canb.auug.org.au>
> Cc: Arvind Sankar <nivedita@alum.mit.edu>
> Cc: linux-riscv@lists.infradead.org
> Cc: clang-built-linux@googlegroups.com
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Nathan Chancellor <natechancellor@gmail.com>
> Cc: Paul Walmsley <paul.walmsley@sifive.com>
> Cc: Palmer Dabbelt <palmer@dabbelt.com>
> Cc: Albert Ou <aou@eecs.berkeley.edu>
> ---
>  arch/riscv/include/asm/vdso/processor.h |    2 ++
>  1 file changed, 2 insertions(+)
>
> --- lnx-510-rc4.orig/arch/riscv/include/asm/vdso/processor.h
> +++ lnx-510-rc4/arch/riscv/include/asm/vdso/processor.h
> @@ -4,6 +4,8 @@
>
>  #ifndef __ASSEMBLY__
>
> +#include <asm/barrier.h>
> +
>  static inline void cpu_relax(void)
>  {
>  #ifdef __riscv_muldiv

Thanks, this is on fixes.
diff mbox series

Patch

--- lnx-510-rc4.orig/arch/riscv/include/asm/vdso/processor.h
+++ lnx-510-rc4/arch/riscv/include/asm/vdso/processor.h
@@ -4,6 +4,8 @@ 
 
 #ifndef __ASSEMBLY__
 
+#include <asm/barrier.h>
+
 static inline void cpu_relax(void)
 {
 #ifdef __riscv_muldiv