[COMMITTED] S390: Fix non-ascii character in fenv.h.

  The comment "isn't" contained a non-ascii character which leads to
an error if compiled with -finput-charset=ascii:
error: failure to convert ascii to UTF-8

This is observable in GCC testsuite:
FAIL: 17_intro/headers/c++1998/charset.cc (test for excess errors)
FAIL: 17_intro/headers/c++2011/charset.cc (test for excess errors)
FAIL: 17_intro/headers/c++2014/charset.cc (test for excess errors)
FAIL: 17_intro/headers/c++2017/charset.cc (test for excess errors)
FAIL: 17_intro/headers/c++2020/charset.cc (test for excess errors)

Also rewrite the comment above.

Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
---
 sysdeps/s390/fpu/bits/fenv.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
  

* Stefan Liebler:

> The comment "isn't" contained a non-ascii character which leads to
> an error if compiled with -finput-charset=ascii:
> error: failure to convert ascii to UTF-8
>
> This is observable in GCC testsuite:
> FAIL: 17_intro/headers/c++1998/charset.cc (test for excess errors)
> FAIL: 17_intro/headers/c++2011/charset.cc (test for excess errors)
> FAIL: 17_intro/headers/c++2014/charset.cc (test for excess errors)
> FAIL: 17_intro/headers/c++2017/charset.cc (test for excess errors)
> FAIL: 17_intro/headers/c++2020/charset.cc (test for excess errors)
>
> Also rewrite the comment above.
>
> Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>

Looks good.

I think we should add -finput-charset=ascii to check-install-headers.
Do you want to send a patch for that?
  

On Feb 12 2020, Stefan Liebler wrote:

> diff --git a/sysdeps/s390/fpu/bits/fenv.h b/sysdeps/s390/fpu/bits/fenv.h
> index 4038f96..e10d025 100644
> --- a/sysdeps/s390/fpu/bits/fenv.h
> +++ b/sysdeps/s390/fpu/bits/fenv.h
> @@ -73,14 +73,14 @@ typedef unsigned int fexcept_t; /* size of fpc */
>  
>  
>  /* Type representing floating-point environment.  This function corresponds
> -   to the layout of the block written by the `fstenv'.  */
> +   to the layout of the block used by fegetenv and fesetenv.  */
>  typedef struct
>  {
>    fexcept_t __fpc;
>    void *__unused;
>    /* The field __unused (formerly __ieee_instruction_pointer) is a relict from
>       commit "Remove PTRACE_PEEKUSER" (87b9b50f0d4b92248905e95a06a13c513dc45e59)
> -     and isn´t used anymore.  */
> +     and isn't used anymore.  */

Perhaps avoid the contraction and write it out as `is not'.

Andreas.
  

On 2/12/20 10:12 AM, Florian Weimer wrote:
> * Stefan Liebler:
> 
>> The comment "isn't" contained a non-ascii character which leads to
>> an error if compiled with -finput-charset=ascii:
>> error: failure to convert ascii to UTF-8
>>
>> This is observable in GCC testsuite:
>> FAIL: 17_intro/headers/c++1998/charset.cc (test for excess errors)
>> FAIL: 17_intro/headers/c++2011/charset.cc (test for excess errors)
>> FAIL: 17_intro/headers/c++2014/charset.cc (test for excess errors)
>> FAIL: 17_intro/headers/c++2017/charset.cc (test for excess errors)
>> FAIL: 17_intro/headers/c++2020/charset.cc (test for excess errors)
>>
>> Also rewrite the comment above.
>>
>> Reported-by: Ulrich Weigand <Ulrich.Weigand@de.ibm.com>
> 
> Looks good.
> 
> I think we should add -finput-charset=ascii to check-install-headers.
> Do you want to send a patch for that?
> 

Sure. Please have a look at this patch:
"[PATCH] Use gcc -finput-charset=ascii for check-installed-headers."
https://www.sourceware.org/ml/libc-alpha/2020-02/msg00505.html

Bye,
Stefan
  

On 2/12/20 10:33 AM, Andreas Schwab wrote:
> On Feb 12 2020, Stefan Liebler wrote:
> 
>> diff --git a/sysdeps/s390/fpu/bits/fenv.h b/sysdeps/s390/fpu/bits/fenv.h
>> index 4038f96..e10d025 100644
>> --- a/sysdeps/s390/fpu/bits/fenv.h
>> +++ b/sysdeps/s390/fpu/bits/fenv.h
>> @@ -73,14 +73,14 @@ typedef unsigned int fexcept_t; /* size of fpc */
>>   
>>   
>>   /* Type representing floating-point environment.  This function corresponds
>> -   to the layout of the block written by the `fstenv'.  */
>> +   to the layout of the block used by fegetenv and fesetenv.  */
>>   typedef struct
>>   {
>>     fexcept_t __fpc;
>>     void *__unused;
>>     /* The field __unused (formerly __ieee_instruction_pointer) is a relict from
>>        commit "Remove PTRACE_PEEKUSER" (87b9b50f0d4b92248905e95a06a13c513dc45e59)
>> -     and isn´t used anymore.  */
>> +     and isn't used anymore.  */
> 
> Perhaps avoid the contraction and write it out as `is not'.
> 
> Andreas.
> 
Yes, I will avoid such contractions in future.
Shall I commit a further patch which for this comment?

Bye,
Stefan