Hi Marcelo,
just finished building 2.4.33-rc3 on my dual-CPU Sun U60 (works
fine BTW). I noticed that smbfs built as a module needs memchr()
since a recent fix, so this one now needs to be exported, which
this patch does. Sources show that the lp driver would need it
too is console on LP is enabled and LP is set as a module (which
seems stupid to me anyway). I've pushed it into -upstream if you
prefer to pull from it.
Overall, 2.4.33-rc3 seems to be OK to me. I don't think that
an additionnal -rc4 would be needed just for this export (Grant
CCed in case he's wishing to do a few more builds, you know
him... :-) ).
Regards,
Willy
>From e3523609bec99d5c607fc00b4f68386d3390fb82 Mon Sep 17 00:00:00 2001
From: Willy Tarreau <[email protected]>
Date: Wed, 2 Aug 2006 23:30:22 +0200
Subject: [PATCH] export memchr() which is used by smbfs and lp driver.
Recently, an smbfs fix added a dependency on memchr() which is
not exported if smbfs is built as a module.
Signed-Off-By: Willy Tarreau <[email protected]>
---
kernel/ksyms.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/kernel/ksyms.c b/kernel/ksyms.c
index d1e66c7..73ad3e9 100644
--- a/kernel/ksyms.c
+++ b/kernel/ksyms.c
@@ -579,6 +579,7 @@ EXPORT_SYMBOL(get_write_access);
EXPORT_SYMBOL(strnicmp);
EXPORT_SYMBOL(strspn);
EXPORT_SYMBOL(strsep);
+EXPORT_SYMBOL(memchr);
#ifdef CONFIG_CRC32
EXPORT_SYMBOL(crc32_le);
--
1.4.1
On Wed, 2 Aug 2006 23:46:08 +0200, Willy Tarreau <[email protected]> wrote:
>Hi Marcelo,
>
>just finished building 2.4.33-rc3 on my dual-CPU Sun U60 (works
>fine BTW). I noticed that smbfs built as a module needs memchr()
>since a recent fix, so this one now needs to be exported, which
>this patch does. Sources show that the lp driver would need it
>too is console on LP is enabled and LP is set as a module (which
>seems stupid to me anyway). I've pushed it into -upstream if you
>prefer to pull from it.
>
>Overall, 2.4.33-rc3 seems to be OK to me. I don't think that
>an additionnal -rc4 would be needed just for this export (Grant
>CCed in case he's wishing to do a few more builds, you know
>him... :-) ).
Just one build, deltree server 'cos it exports some shares to 'doze
boxen. Seems to be okay without the patch but...
Makes no difference to my limited usage here. Didn't break ;)
<http://bugsplatter.mine.nu/test/boxen/deltree/> *-rc3a
uptime 21 mins...
Cheers,
Grant.
On Thu, Aug 03, 2006 at 11:21:51AM +1000, Grant Coady wrote:
> On Wed, 2 Aug 2006 23:46:08 +0200, Willy Tarreau <[email protected]> wrote:
>
> >Hi Marcelo,
> >
> >just finished building 2.4.33-rc3 on my dual-CPU Sun U60 (works
> >fine BTW). I noticed that smbfs built as a module needs memchr()
> >since a recent fix, so this one now needs to be exported, which
> >this patch does. Sources show that the lp driver would need it
> >too is console on LP is enabled and LP is set as a module (which
> >seems stupid to me anyway). I've pushed it into -upstream if you
> >prefer to pull from it.
> >
> >Overall, 2.4.33-rc3 seems to be OK to me. I don't think that
> >an additionnal -rc4 would be needed just for this export (Grant
> >CCed in case he's wishing to do a few more builds, you know
> >him... :-) ).
>
> Just one build, deltree server 'cos it exports some shares to 'doze
> boxen. Seems to be okay without the patch but...
>
> Makes no difference to my limited usage here. Didn't break ;)
Interesting, I wonder whether memchr() may be a gcc builtin on some
archs, explaining why it works without the patch on your machine.
> <http://bugsplatter.mine.nu/test/boxen/deltree/> *-rc3a
>
> uptime 21 mins...
Thanks very much, that's exactly what I needed.
> Cheers,
> Grant.
Cheers,
Willy
On Thu, 3 Aug 2006 04:40:37 +0200, Willy Tarreau <[email protected]> wrote:
>On Thu, Aug 03, 2006 at 11:21:51AM +1000, Grant Coady wrote:
>> On Wed, 2 Aug 2006 23:46:08 +0200, Willy Tarreau <[email protected]> wrote:
>>
>> >Hi Marcelo,
>> >
>> >just finished building 2.4.33-rc3 on my dual-CPU Sun U60 (works
>> >fine BTW). I noticed that smbfs built as a module needs memchr()
>> >since a recent fix, so this one now needs to be exported, which
>> >this patch does. Sources show that the lp driver would need it
>> >too is console on LP is enabled and LP is set as a module (which
>> >seems stupid to me anyway). I've pushed it into -upstream if you
>> >prefer to pull from it.
>> >
>> >Overall, 2.4.33-rc3 seems to be OK to me. I don't think that
>> >an additionnal -rc4 would be needed just for this export (Grant
>> >CCed in case he's wishing to do a few more builds, you know
>> >him... :-) ).
>>
>> Just one build, deltree server 'cos it exports some shares to 'doze
>> boxen. Seems to be okay without the patch but...
>>
>> Makes no difference to my limited usage here. Didn't break ;)
>
>Interesting, I wonder whether memchr() may be a gcc builtin on some
>archs, explaining why it works without the patch on your machine.
Compile host is Slackware-10.2, and:
~$ gcc --version
gcc (GCC) 3.3.6
Grant.
On Thu, Aug 03, 2006 at 01:26:53PM +1000, Grant Coady wrote:
> On Thu, 3 Aug 2006 04:40:37 +0200, Willy Tarreau <[email protected]> wrote:
>
> >On Thu, Aug 03, 2006 at 11:21:51AM +1000, Grant Coady wrote:
> >> On Wed, 2 Aug 2006 23:46:08 +0200, Willy Tarreau <[email protected]> wrote:
> >>
> >> >Hi Marcelo,
> >> >
> >> >just finished building 2.4.33-rc3 on my dual-CPU Sun U60 (works
> >> >fine BTW). I noticed that smbfs built as a module needs memchr()
> >> >since a recent fix, so this one now needs to be exported, which
> >> >this patch does. Sources show that the lp driver would need it
> >> >too is console on LP is enabled and LP is set as a module (which
> >> >seems stupid to me anyway). I've pushed it into -upstream if you
> >> >prefer to pull from it.
> >> >
> >> >Overall, 2.4.33-rc3 seems to be OK to me. I don't think that
> >> >an additionnal -rc4 would be needed just for this export (Grant
> >> >CCed in case he's wishing to do a few more builds, you know
> >> >him... :-) ).
> >>
> >> Just one build, deltree server 'cos it exports some shares to 'doze
> >> boxen. Seems to be okay without the patch but...
> >>
> >> Makes no difference to my limited usage here. Didn't break ;)
> >
> >Interesting, I wonder whether memchr() may be a gcc builtin on some
> >archs, explaining why it works without the patch on your machine.
>
> Compile host is Slackware-10.2, and:
> ~$ gcc --version
> gcc (GCC) 3.3.6
OK, I used gcc-3.3.5 on sparc64.
> Grant.
Willy