[Cooker] Re: kernel and gcc3.2, a good match?

[Juan Quintela]

> "lonnie" == Lonnie Borntreger <[EMAIL PROTECTED]> writes:

lonnie> On Wed, 2002-11-20 at 16:00, Edward Tandi wrote:
>> On Wed, 2002-11-20 at 21:42, Lonnie Borntreger wrote:
>> > On Wed, 2002-11-20 at 14:55, Edward Tandi wrote:
>> > > I too had problems with gcc3.2 on my twin Athlon MP system. The kernel
>> > > warns about:
>> > > 
>> > > kernel: Advanced speculative caching feature present
>> > > kernel: Disabling advanced speculative caching
>> > 
>> > I'm running kernel-2.4.19-16 on a dual XP box, compiled with gcc3.2 with
>> > total stability.  Of course, I always grab the kernel src.rpm and
>> > recompile it for athlon.
>> 
>> This is good to know. Of course, the problems might be driver specific.
>> What motherboard are you using? Any additional I/O cards?

lonnie> Details:
lonnie> ASUS A7M266-D
lonnie> on board ATA-100 controller
lonnie> on board CM8738 audio (no mic in - alsa driver related)
lonnie> USB2.0 card (no USB devices)
lonnie> nVidia TNT2 AGP (nVidia drivers)
lonnie> Hauppage WinTV/Radio/IR

Try:
a) without nvidia drivers, or
b) with acpi=off

It appears that there is some problem with nvidia drivers and acpi
kernel :(

Later, Juan.


lonnie> Any issues I have/had are due to recently switching from self compiled
lonnie> kernel.org kernels with OSS sound, and no devfs.  The devfs is still
lonnie> giving me issues with trying to get some things working, but the kernel
lonnie> itself is perfect.

lonnie> TTFN, 
lonnie> Lonnie Borntreger




-- 
In theory, practice and theory are the same, but in practice they 
are different -- Larry McVoy

[Cooker] Re: kernel and gcc3.2, a good match?

[Juan Quintela]

> "edward" == Edward Tandi <[EMAIL PROTECTED]> writes:

edward> I too had problems with gcc3.2 on my twin Athlon MP system. The kernel
edward> warns about:

edward> kernel: Advanced speculative caching feature present
edward> kernel: Disabling advanced speculative caching

This is not a gcc bug.  It is a kernel bug for Athlons, and this is
just showing you that it is applying a workaround for it :)

Believe me, you don't want to know the details :)

edward> So, it might just work now, but at the time, it crashed for me under I/O
edward> load and sometimes while using X.

edward> Like yourself, I went back to using an earlier kernel. I still use
edward> 2.4.18-pre9 compiled with gcc 2.96 and it is solid. It's my main server
edward> and I worry about newer kernels, so I don't touch.

edward> I know kernel developers have been improving things by porting bits of
edward> the kernel with reported problems to work with gcc 3.2, but I still
edward> think it's not the official build compiler. I look forward to when it is
edward> because gcc 3.2 has some good performance optimisations.

Humm, I haven't had any problems with gcc since gcc 3.1.1 or something
like that.  Could you told me if you still have problems?

Later,, Juan.

-- 
In theory, practice and theory are the same, but in practice they 
are different -- Larry McVoy