Hi, 

Isn't necessary to update general.ent?

Jean-Philippe MENGUAL


Le jeudi 11 juin 2009 à 23:25 -0600, [email protected] a
écrit :
> Author: bryan
> Date: 2009-06-11 23:25:14 -0600 (Thu, 11 Jun 2009)
> New Revision: 8944
> 
> Modified:
>    trunk/BOOK/chapter01/changelog.xml
>    trunk/BOOK/chapter05/glibc.xml
>    trunk/BOOK/chapter06/glibc.xml
> Log:
> Add a sed to glibc to fix the constants passed to futex(2) in 32-bit builds.  
> Seems to fix testsuite hangs (the tst-mutex9 process never exits, for 
> instance).
> 
> This change is in upstream git, so this can be reverted once we upgrade glibc.
> 
> 
> 
> Modified: trunk/BOOK/chapter01/changelog.xml
> ===================================================================
> --- trunk/BOOK/chapter01/changelog.xml        2009-06-12 05:20:23 UTC (rev 
> 8943)
> +++ trunk/BOOK/chapter01/changelog.xml        2009-06-12 05:25:14 UTC (rev 
> 8944)
> @@ -38,6 +38,17 @@
>  -->
>  
>      <listitem>
> +      <para>2009-06-11</para>
> +      <itemizedlist>
> +        <listitem>
> +          <para>[bryan] - Add a sed to glibc to fix the constants passed
> +          to futex(2) in 32-bit builds with certain --enable-kernel values,
> +          including our 2.6.18.</para>
> +        </listitem>
> +      </itemizedlist>
> +    </listitem>
> +
> +    <listitem>
>        <para>2009-06-06</para>
>        <itemizedlist>
>          <listitem>
> 
> Modified: trunk/BOOK/chapter05/glibc.xml
> ===================================================================
> --- trunk/BOOK/chapter05/glibc.xml    2009-06-12 05:20:23 UTC (rev 8943)
> +++ trunk/BOOK/chapter05/glibc.xml    2009-06-12 05:25:14 UTC (rev 8944)
> @@ -43,6 +43,16 @@
>    <sect2 role="installation">
>      <title>Installation of Glibc</title>
>  
> +    <!-- This sed can be removed for any glibc newer than 2.10.1 (along with 
> the
> +         matching sed in chapter 6).  See glibc bug 10262. -->
> +    <para>Fix an error in the constants that get passed to the futex system
> +    call in some cases, causing certain pthread_mutex operations to 
> fail:</para>
> +
> +<screen><userinput remap="pre">cp 
> nptl/sysdeps/unix/sysv/linux/i386/i486/lowlevellock.S{,.orig}
> +sed -e 's/FUTEX_WAIT\( | FUTEX_CLOCK_REALTIME, reg\)/FUTEX_WAIT_BITSET\1/' \
> +        &lt;nptl/sysdeps/unix/sysv/linux/i386/i486/lowlevellock.S.orig \
> +        
> &gt;nptl/sysdeps/unix/sysv/linux/i386/i486/lowlevellock.S</userinput></screen>
> +
>      <para>The Glibc documentation recommends building Glibc outside of the 
> source
>      directory in a dedicated build directory:</para>
>  
> 
> Modified: trunk/BOOK/chapter06/glibc.xml
> ===================================================================
> --- trunk/BOOK/chapter06/glibc.xml    2009-06-12 05:20:23 UTC (rev 8943)
> +++ trunk/BOOK/chapter06/glibc.xml    2009-06-12 05:25:14 UTC (rev 8944)
> @@ -108,6 +108,14 @@
>  sed -i '/tst-fgetws-ENV/ a\
>  tst-fgetwc-ENV = LOCPATH=$(common-objpfx)localedata' 
> libio/Makefile</userinput></screen>
>  
> +    <!-- This sed can be removed for any glibc newer than 2.10.1 (along with 
> the
> +         matching sed in chapter 5).  See glibc bug 10262. -->
> +    <para>Fix an error in the constants that get passed to the futex system
> +    call in some cases, causing certain pthread_mutex operations to 
> fail:</para>
> +
> +<screen><userinput remap="pre">sed -i -e 's/FUTEX_WAIT\( | 
> FUTEX_CLOCK_REALTIME, reg\)/FUTEX_WAIT_BITSET\1/' \
> +        
> nptl/sysdeps/unix/sysv/linux/i386/i486/lowlevellock.S</userinput></screen>
> +
>      <para>The Glibc documentation recommends building Glibc outside of the 
> source
>      directory in a dedicated build directory:</para>
>  
> 

-- 
http://linuxfromscratch.org/mailman/listinfo/lfs-book
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

Reply via email to