unofficial mirror of libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: Stefan Liebler <stli@linux.ibm.com>
To: GNU C Library <libc-alpha@sourceware.org>, libc-stable@sourceware.org
Cc: "Carlos O'Donell" <carlos@redhat.com>
Subject: [2.29 COMMITTED] Fix output of LD_SHOW_AUXV=1.
Date: Wed, 13 Mar 2019 10:54:29 +0100	[thread overview]
Message-ID: <26473622-44fc-b2ae-6132-e7d475190f95@linux.ibm.com> (raw)
In-Reply-To: <5b73a948-5209-14b0-61d1-aca82c99e192@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 4411 bytes --]

On 3/12/19 9:11 PM, Carlos O'Donell wrote:
> On 3/12/19 11:06 AM, Stefan Liebler wrote:
>> commit 988376a08d6d36ea321203b9c039db4b266609ae
>> Author: Stefan Liebler <stli@linux.ibm.com>
>> Date:   Fri Mar 8 11:18:53 2019 +0100
>>
>>      Fix output of LD_SHOW_AUXV=1.
>>      
>>      Starting with commit 1616d034b61622836d3a36af53dcfca7624c844e
>>      the output was corrupted on some platforms as _dl_procinfo
>>      was called for every auxv entry and on some architectures like s390
>>      all entries were represented as "AT_HWCAP".
>>      
>>      This patch is removing the condition and let _dl_procinfo decide if
>>      an entry is printed in a platform specific or generic way.
>>      This patch also adjusts all _dl_procinfo implementations which assumed
>>      that they are only called for AT_HWCAP or AT_HWCAP2. They are now just
>>      returning a non-zero-value for entries which are not handled platform
>>      specifc.
>>      
>>      ChangeLog:
>>      
>>              * elf/dl-sysdep.c (_dl_show_auxv): Remove condition and always
>>              call _dl_procinfo.
>>              * sysdeps/unix/sysv/linux/s390/dl-procinfo.h (_dl_procinfo):
>>              Ignore types other than AT_HWCAP.
>>              * sysdeps/sparc/dl-procinfo.h (_dl_procinfo): Likewise.
>>              * sysdeps/unix/sysv/linux/i386/dl-procinfo.h (_dl_procinfo):
>>              Likewise.
> 
> Thanks for the cleanup, this makes it really clear how this is intended to work.
> 
> OK for master if you...
> - Fix ARM and POWERPC backend _dl_procinfo to remove "/* This should not happen.  */"
>    comment in dl-procinfo.c. It *does* happen now because we pass all options to
>    the machine version first.
> 
> Reviewed-by: Carlos O'Donell <carlos@redhat.com>
> 
Okay. I've adjusted those comments and committed the patch to master and 
backported it to glibc release branch 2.29.

Thanks,
Stefan

>> diff --git a/elf/dl-sysdep.c b/elf/dl-sysdep.c
>> index 5f6c679a3f..5d19b100b2 100644
>> --- a/elf/dl-sysdep.c
>> +++ b/elf/dl-sysdep.c
>> @@ -328,14 +328,9 @@ _dl_show_auxv (void)
>>         assert (AT_NULL == 0);
>>         assert (AT_IGNORE == 1);
>>   
>> -      if (av->a_type == AT_HWCAP || av->a_type == AT_HWCAP2
>> -	  || AT_L1I_CACHEGEOMETRY || AT_L1D_CACHEGEOMETRY
>> -	  || AT_L2_CACHEGEOMETRY || AT_L3_CACHEGEOMETRY)
>> -	{
>> -	  /* These are handled in a special way per platform.  */
>> -	  if (_dl_procinfo (av->a_type, av->a_un.a_val) == 0)
>> -	    continue;
>> -	}
>> +      /* Some entries are handled in a special way per platform.  */
>> +      if (_dl_procinfo (av->a_type, av->a_un.a_val) == 0)
>> +	continue;
> 
> OK.
> 
> OK. ARM backend returns -1 with default handling.
> 
> OK. POWERPC backend return -1 with default handling.
> 
>>   
>>         if (idx < sizeof (auxvars) / sizeof (auxvars[0])
>>   	  && auxvars[idx].form != unknown)
>> diff --git a/sysdeps/sparc/dl-procinfo.h b/sysdeps/sparc/dl-procinfo.h
>> index 282b8c5117..cc2687e99b 100644
>> --- a/sysdeps/sparc/dl-procinfo.h
>> +++ b/sysdeps/sparc/dl-procinfo.h
>> @@ -32,7 +32,7 @@ _dl_procinfo (unsigned int type, unsigned long int word)
>>     int i;
>>   
>>     /* Fallback to unknown output mechanism.  */
>> -  if (type == AT_HWCAP2)
>> +  if (type != AT_HWCAP)
> 
> OK.
> 
>>       return -1;
>>   
>>     _dl_printf ("AT_HWCAP:   ");
>> diff --git a/sysdeps/unix/sysv/linux/i386/dl-procinfo.h b/sysdeps/unix/sysv/linux/i386/dl-procinfo.h
>> index 22b43431bc..3aef14c6c1 100644
>> --- a/sysdeps/unix/sysv/linux/i386/dl-procinfo.h
>> +++ b/sysdeps/unix/sysv/linux/i386/dl-procinfo.h
>> @@ -31,7 +31,7 @@ _dl_procinfo (unsigned int type, unsigned long int word)
>>     int i;
>>   
>>     /* Fallback to unknown output mechanism.  */
>> -  if (type == AT_HWCAP2)
>> +  if (type != AT_HWCAP)
> 
> OK.
> 
>>       return -1;
>>   
>>     _dl_printf ("AT_HWCAP:   ");
>> diff --git a/sysdeps/unix/sysv/linux/s390/dl-procinfo.h b/sysdeps/unix/sysv/linux/s390/dl-procinfo.h
>> index 19329a335b..16739fd6cd 100644
>> --- a/sysdeps/unix/sysv/linux/s390/dl-procinfo.h
>> +++ b/sysdeps/unix/sysv/linux/s390/dl-procinfo.h
>> @@ -33,7 +33,7 @@ _dl_procinfo (unsigned int type, unsigned long int word)
>>     int i;
>>   
>>     /* Fallback to unknown output mechanism.  */
>> -  if (type == AT_HWCAP2)
>> +  if (type != AT_HWCAP)
> 
> OK.
> 
>>       return -1;
>>   
>>     _dl_printf ("AT_HWCAP:   ");
> 
> 


[-- Attachment #2: 20190313_ldshowauxv.patch --]
[-- Type: text/x-patch, Size: 4692 bytes --]

commit daede01d78492f2f7b640966a2e60ff409be677a
Author: Stefan Liebler <stli@linux.ibm.com>
Date:   Fri Mar 8 11:18:53 2019 +0100

    Fix output of LD_SHOW_AUXV=1.
    
    Starting with commit 1616d034b61622836d3a36af53dcfca7624c844e
    the output was corrupted on some platforms as _dl_procinfo
    was called for every auxv entry and on some architectures like s390
    all entries were represented as "AT_HWCAP".
    
    This patch is removing the condition and let _dl_procinfo decide if
    an entry is printed in a platform specific or generic way.
    This patch also adjusts all _dl_procinfo implementations which assumed
    that they are only called for AT_HWCAP or AT_HWCAP2. They are now just
    returning a non-zero-value for entries which are not handled platform
    specifc.
    
    ChangeLog:
    
            * elf/dl-sysdep.c (_dl_show_auxv): Remove condition and always
            call _dl_procinfo.
            * sysdeps/unix/sysv/linux/s390/dl-procinfo.h (_dl_procinfo):
            Ignore types other than AT_HWCAP.
            * sysdeps/sparc/dl-procinfo.h (_dl_procinfo): Likewise.
            * sysdeps/unix/sysv/linux/i386/dl-procinfo.h (_dl_procinfo):
            Likewise.
            * sysdeps/powerpc/dl-procinfo.h (_dl_procinfo): Adjust comment
            in the case of falling back to generic output mechanism.
            * sysdeps/unix/sysv/linux/arm/dl-procinfo.h (_dl_procinfo):
            Likewise.

diff --git a/elf/dl-sysdep.c b/elf/dl-sysdep.c
index 5f6c679a3f..5d19b100b2 100644
--- a/elf/dl-sysdep.c
+++ b/elf/dl-sysdep.c
@@ -328,14 +328,9 @@ _dl_show_auxv (void)
       assert (AT_NULL == 0);
       assert (AT_IGNORE == 1);
 
-      if (av->a_type == AT_HWCAP || av->a_type == AT_HWCAP2
-	  || AT_L1I_CACHEGEOMETRY || AT_L1D_CACHEGEOMETRY
-	  || AT_L2_CACHEGEOMETRY || AT_L3_CACHEGEOMETRY)
-	{
-	  /* These are handled in a special way per platform.  */
-	  if (_dl_procinfo (av->a_type, av->a_un.a_val) == 0)
-	    continue;
-	}
+      /* Some entries are handled in a special way per platform.  */
+      if (_dl_procinfo (av->a_type, av->a_un.a_val) == 0)
+	continue;
 
       if (idx < sizeof (auxvars) / sizeof (auxvars[0])
 	  && auxvars[idx].form != unknown)
diff --git a/sysdeps/powerpc/dl-procinfo.h b/sysdeps/powerpc/dl-procinfo.h
index f542f7318f..dfc3b33a72 100644
--- a/sysdeps/powerpc/dl-procinfo.h
+++ b/sysdeps/powerpc/dl-procinfo.h
@@ -225,7 +225,7 @@ _dl_procinfo (unsigned int type, unsigned long int word)
 	break;
       }
     default:
-      /* This should not happen.  */
+      /* Fallback to generic output mechanism.  */
       return -1;
     }
    _dl_printf ("\n");
diff --git a/sysdeps/sparc/dl-procinfo.h b/sysdeps/sparc/dl-procinfo.h
index 282b8c5117..64ee267fc7 100644
--- a/sysdeps/sparc/dl-procinfo.h
+++ b/sysdeps/sparc/dl-procinfo.h
@@ -31,8 +31,8 @@ _dl_procinfo (unsigned int type, unsigned long int word)
 {
   int i;
 
-  /* Fallback to unknown output mechanism.  */
-  if (type == AT_HWCAP2)
+  /* Fallback to generic output mechanism.  */
+  if (type != AT_HWCAP)
     return -1;
 
   _dl_printf ("AT_HWCAP:   ");
diff --git a/sysdeps/unix/sysv/linux/arm/dl-procinfo.h b/sysdeps/unix/sysv/linux/arm/dl-procinfo.h
index 66c00297b7..05c62c8687 100644
--- a/sysdeps/unix/sysv/linux/arm/dl-procinfo.h
+++ b/sysdeps/unix/sysv/linux/arm/dl-procinfo.h
@@ -67,7 +67,7 @@ _dl_procinfo (unsigned int type, unsigned long int word)
 	break;
       }
     default:
-      /* This should not happen.  */
+      /* Fallback to generic output mechanism.  */
       return -1;
     }
   _dl_printf ("\n");
diff --git a/sysdeps/unix/sysv/linux/i386/dl-procinfo.h b/sysdeps/unix/sysv/linux/i386/dl-procinfo.h
index 22b43431bc..0585cdaa9c 100644
--- a/sysdeps/unix/sysv/linux/i386/dl-procinfo.h
+++ b/sysdeps/unix/sysv/linux/i386/dl-procinfo.h
@@ -30,8 +30,8 @@ _dl_procinfo (unsigned int type, unsigned long int word)
      in the kernel sources.  */
   int i;
 
-  /* Fallback to unknown output mechanism.  */
-  if (type == AT_HWCAP2)
+  /* Fallback to generic output mechanism.  */
+  if (type != AT_HWCAP)
     return -1;
 
   _dl_printf ("AT_HWCAP:   ");
diff --git a/sysdeps/unix/sysv/linux/s390/dl-procinfo.h b/sysdeps/unix/sysv/linux/s390/dl-procinfo.h
index 19329a335b..d67fde368f 100644
--- a/sysdeps/unix/sysv/linux/s390/dl-procinfo.h
+++ b/sysdeps/unix/sysv/linux/s390/dl-procinfo.h
@@ -32,8 +32,8 @@ _dl_procinfo (unsigned int type, unsigned long int word)
      in the kernel sources.  */
   int i;
 
-  /* Fallback to unknown output mechanism.  */
-  if (type == AT_HWCAP2)
+  /* Fallback to generic output mechanism.  */
+  if (type != AT_HWCAP)
     return -1;
 
   _dl_printf ("AT_HWCAP:   ");

      reply	other threads:[~2019-03-13  9:54 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-03-08 12:52 [PATCH] Fix output of LD_SHOW_AUXV=1 Stefan Liebler
2019-03-08 14:51 ` Carlos O'Donell
2019-03-12 15:06   ` Stefan Liebler
2019-03-12 20:11     ` Carlos O'Donell
2019-03-13  9:54       ` Stefan Liebler [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/libc/involved.html

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=26473622-44fc-b2ae-6132-e7d475190f95@linux.ibm.com \
    --to=stli@linux.ibm.com \
    --cc=carlos@redhat.com \
    --cc=libc-alpha@sourceware.org \
    --cc=libc-stable@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).