lkml.org 
[lkml]   [2003]   [Feb]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: Unexpected lock during "Calibrating delay loop" and failure to compile without "HighRes"
Isabelle, Francois wrote:> Compilation with high-res turned off:
>
>
> add #define do_gettimeoffset() do_slow_gettimeoffset() at line 283 in
> arch/i386/kernel/time.c

No, I think not. That messes up other logic. This is what I think it
should be (note the below is the idea, a correct patch is attached):
xx@@ -869,7 +869,7 @@
* and just enable this for the next intel chips ?
*/
x86_udelay_tsc = 1;
-#ifndef do_gettimeoffset
+#if ! defined( do_gettimeoffset) && defined(CONFIG_X86_TSC)
do_gettimeoffset = do_fast_gettimeoffset;
#endif
/*

Thanks for the bug report.

>
> their was a
> "static unsigned long (*do_gettimeoffset)(void) = do_slow_gettimeoffset;"
> instead of the #define found for other configuration
>
> at line 875
>
> #ifndef do_gettimeoffset
> do_gettimeoffset = do_fast_gettimeoffset;
> #endif
>
> would be compiled even when the CONFIG_X86_TSC is off.
>
>
>
>>-----Original Message-----
>>From: Isabelle, Francois [mailto:Francois.Isabelle@ca.kontron.com]
>>Sent: 3 février, 2003 10:07
>>To: Isabelle, Francois;
>>high-res-timers-discourse@lists.sourceforge.net
>>Cc: linux-kernel@vger.kernel.org
>>Subject: RE: Errata : Unexpected lock during "Calibrating delay loop"
>>and failure to co mpile without "HighRes"
>>
>>
>>Please Read:
>> to build withOUT High-Res, the kernel won't build
>>
>>
>>>-----Original Message-----
>>>From: Isabelle, Francois [mailto:Francois.Isabelle@ca.kontron.com]
>>>Sent: 3 février, 2003 09:47
>>>To: high-res-timers-discourse@lists.sourceforge.net
>>>Cc: linux-kernel@vger.kernel.org
>>>Subject: Unexpected lock during "Calibrating delay loop" and
>>>failure to
>>>co mpile without "HighRes"
>>>
>>>
>>>Hi,
>>> I'm trying to integrate some tools on a 486-powered cpu
>>>board, I don't
>>>really need "High Resolution Timers", but one of the tools
>>>would really make
>>>good use of the POSIX API you implemented. I've patch kernel
>>>2.4.20 with the
>>>latest 2.4.20-1.0 hrtimers.
>>>
>>>Here comes the trouble.
>>>
>>>- Trying to build with High-Res, the kernel won't build
>>>
>>>time.c: In function `time_init':
>>>time.c:873: `do_fast_gettimeoffset' undeclared (first use in
>>>this function)
>>>time.c:873: (Each undeclared identifier is reported only once
>>>time.c:873: for each function it appears in.)
>>>make[1]: *** [time.o] Error 1
>>>make[1]: Leaving directory `/usr/src/linux-2.4.20/arch/i386/kernel'
>>>make: *** [_dir_arch/i386/kernel] Error 2
>>>
>>>seems like it should try to link "do_slow_gettimeoffset"
>>>instead since 486
>>>does not handle TSC, (I'll have to check that..)
>>>
>>>
>>>- Trying to boot with "PIT-based" high-res support, the
>>>kernel lock during
>>>calibration "Calibrating delay loop".
>>> Same occurs with IOAPIC and TSC ...
>>>
>>>
>>>If you have any hint, I'll be glad to hear it.
>>>
>>>Thanks
>>>
>>>
>>>Frank
>>>
>>>
>>>
>>>-------------------------------------------------------
>>>This SF.NET email is sponsored by:
>>>SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
>>>http://www.vasoftware.com
>>>to unsubscribe:
>>
>>http://lists.sourceforge.net/lists/listinfo/high-res-timers-discourse
>>High-res-timers-discourse mailing list
>>High-res-timers-discourse@lists.sourceforge.net
>>https://lists.sourceforge.net/lists/listinfo/high-res-timers-discourse
>>
>>
>>-------------------------------------------------------
>>This SF.NET email is sponsored by:
>>SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
>>http://www.vasoftware.com
>>to unsubscribe:
>>http://lists.sourceforge.net/lists/listinfo/high-res-timers-discourse
>>High-res-timers-discourse mailing list
>>High-res-timers-discourse@lists.sourceforge.net
>>https://lists.sourceforge.net/lists/listinfo/high-res-timers-discourse
>>
>
>
>
> -------------------------------------------------------
> This SF.NET email is sponsored by:
> SourceForge Enterprise Edition + IBM + LinuxWorld http://www.vasoftware.com
> to unsubscribe: http://lists.sourceforge.net/lists/listinfo/high-res-timers-discourse
> High-res-timers-discourse mailing list
> High-res-timers-discourse@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/high-res-timers-discourse
>
>



--
George Anzinger george@mvista.com
High-res-timers: http://sourceforge.net/projects/high-res-timers/
Preemption patch: http://www.kernel.org/pub/linux/kernel/people/rml
--- linux-2.4.20-hrt/arch/i386/kernel/time.c~ 2003-01-08 22:51:06.000000000 -0800
+++ linux/arch/i386/kernel/time.c 2003-02-03 14:22:31.000000000 -0800
@@ -869,7 +869,7 @@
* and just enable this for the next intel chips ?
*/
x86_udelay_tsc = 1;
-#ifndef do_gettimeoffset
+#if ! defined( do_gettimeoffset) && defined(CONFIG_X86_TSC)
do_gettimeoffset = do_fast_gettimeoffset;
#endif
/*
\
 
 \ /
  Last update: 2005-03-22 13:32    [W:0.131 / U:0.372 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site