Could it be some buffer is some times allocated (word/page/whatever)
aligned and sometimes not?? Tends to make a big difference on Intel..
Mike
Jon Pennington wrote:
>
> On Wed, Apr 04, 2001 at 06:30:51PM -0600, Nicholas Leippe wrote:
>
> > where it varies across the entire range of possibil
On Thursday 05 April 2001 03:08 am, Mike Lincoln wrote:
> Could it be some buffer is some times allocated (word/page/whatever)
> aligned and sometimes not?? Tends to make a big difference on Intel..
>
> Mike
Well, I thought about this. But looking at the source (it's pretty
straight forward) th
I _did_ see (unless i was up waaay too late again) a SiS DRM driver in the
last DRI code I checked out, and also the corresponding SiS driver in the
X tree ...
I briefly looked inside and it said that it supports the SiS630/540 and
a couple other models...
Did i see things right? And why d
On Wed, Apr 04, 2001 at 01:44:27PM -0500, [EMAIL PROTECTED] wrote:
>
> I _did_ see (unless i was up waaay too late again) a SiS DRM driver in the
> last DRI code I checked out, and also the corresponding SiS driver in the
> X tree ...
> I briefly looked inside and it said that it supports the S
--snip---
>>
>> Maybe I should send it to Gareth with that beer I owe him. :)
>Excuse me, but WHAT's the problem here ??
>I'm using a Hauppauge TV card (BT878 based) together with the DRI and >it
>run PERFECTLY well !!!
>This is using the tdfx driver on a V5.
yeah, same goes for me, bt878 card
Thanks for the quick response Alan,
>SiS themselves dropped us the code quite some time ago and have not
>really updated the driver since the initial drop.
aha! that explains the fact that i haven't seen word about it on the lists..
>None of the active DRI developers have access to such c
On Thu, Apr 05, 2001 at 01:09:11PM +0200, Stefan Lange wrote:
> >> Maybe I should send it to Gareth with that beer I owe him. :)
> >Excuse me, but WHAT's the problem here ??
> >I'm using a Hauppauge TV card (BT878 based) together with the DRI and >it
> >run PERFECTLY well !!!
>
> >This is using t
On Wed, Apr 04, 2001 at 02:47:09PM -0500, [EMAIL PROTECTED] wrote:
>
> Thanks for the quick response Alan,
>
No problem.
> PS. would you know if there's anything funky i need to do to get the kernel
> DRM module going? I didn't see it being spat out of the main Makefile, even
> though there
I've used my WinTV card with a Radeon, i810, V3 and V5, and G400. I
haven't had any problems with those combinations. The only time I've had
an issue was when I used a Rainbow Runner G-Series card with the G400.
If I had the RRG modules loaded, all GL applications had some form of
corruption
Nicholas Leippe wrote:
>
> On Wednesday 04 April 2001 06:07 pm, Brian Paul wrote:
> > This isn't really unexpected. Benchmarking any application on
> > a time-sharing operating system can reveal some variation in
> > performance. You never know when another process or the kernel
> > itself will
> From: Nicholas Leippe <[EMAIL PROTECTED]>
> On Thursday 05 April 2001 03:08 am, Mike Lincoln wrote:
> > Could it be some buffer is some times allocated (word/page/whatever)
> > aligned and sometimes not?? Tends to make a big difference on Intel..
> Well, I thought about this. But looking at
On Thursday 05 April 2001 08:25 am, Jeff Epler wrote:
>
> It's likely due to Linux' lack of page coloring.
>
> Let's see if I can get the explanation right ... The L2 cache is typically
> physically mapped, and so for some value N the addresses A and A+N alias
> to the same cache line. N is some
>>> I have a generic Brooktree 878HKF card for anyone serious about getting DRI
>>> and TV working together.
> I'm using a Hauppauge TV card (BT878 based) together with the DRI and it
> run PERFECTLY well !!!
>
> This is using the tdfx driver on a V5.
> yeah, same goes for me, bt878 card and v3
On Wednesday 04 April 2001 18:51, you wrote:
> This may be opening a pointless can of worms,
> but, I'm really curious as to how to explain
> the behavior I'm seeing. Each time I run
> gears, it is perfectly consistent in it's
> speed (+/- .5 FPS) for that instance. But,
> each time I run gears,
Bill Crawford wrote:
>
> The variation would be in terms of cache usage, e.g. sometimes you
> get buffers that don't collide at all in the cache, other times two
> or more pages happen to collide in the cache and therefore they are
> competing for the same cache lines (leading to more TLB misses
Nicholas Leippe wrote:
>
> Ah yes! I totally forgot about cache contention. This would definitely
> explain it.
In general, yes. However, as most of the traffic is being written out
to AGP memory this is less important. Caching effects in the rest of
the driver may result in the differences,
Gareth Hughes wrote:
>
> In general, this is true. However:
>
> 1) All buffers (malloc()'d and DMA) are nicely aligned to at least a
> cacheline boundary.
>
> 2) DMA buffers reside in uncached AGP memory. Hence, caching plays no
> part in the timing of the app.
Time for a coffee. Please ign
Gareth Hughes wrote:
> In general, this is true. However:
> 1) All buffers (malloc()'d and DMA) are nicely aligned to at least a
> cacheline boundary.
malloc() will not always be allocated contiguous physical pages, so
there is the "colouring" problem ... we might get several pages that
are c
Bill Crawford wrote:
>
> Gareth Hughes wrote:
>
> > In general, this is true. However:
>
> > 1) All buffers (malloc()'d and DMA) are nicely aligned to at least a
> > cacheline boundary.
>
> malloc() will not always be allocated contiguous physical pages, so
> there is the "colouring" problem
Gareth Hughes wrote:
> Time for a coffee. Please ignore the last sentence, I don't know what
> I'm talking about...
I know the feeling ... time I went home for coffee (and sleep :·)
> -- Gareth
While I think of it, I'd like to have a poke at the Rage Pro code
that was on one of the branches
Alan Hourihane wrote:
>
> On Thu, Apr 05, 2001 at 07:41:34PM +0100, Bill Crawford wrote:
> > Gareth Hughes wrote:
> >
> > > Time for a coffee. Please ignore the last sentence, I don't know what
> > > I'm talking about...
> >
> > I know the feeling ... time I went home for coffee (and sleep :·)
On Thu, Apr 05, 2001 at 07:41:34PM +0100, Bill Crawford wrote:
> Gareth Hughes wrote:
>
> > Time for a coffee. Please ignore the last sentence, I don't know what
> > I'm talking about...
>
> I know the feeling ... time I went home for coffee (and sleep :·)
>
Huh ? It's early.. It's only 8:00p
Alan Hourihane wrote:
>
> On Thu, Apr 05, 2001 at 07:41:34PM +0100, Bill Crawford wrote:
> > Gareth Hughes wrote:
> >
> > > Time for a coffee. Please ignore the last sentence, I don't know what
> > > I'm talking about...
> >
> > I know the feeling ... time I went home for coffee (and sleep :·)
Hello, all. I have been attempting to build the mesa-3-5-branch for
about a week now,
and am not having too much success :). I had the problem where matypes.h
was non-existant, but
I _did_ have Mesa(trunk) checked out and built. Now when I build DRI I
get errors while building
gen_matypes.c. Many
I just updated my tree and got this at the top of host.def:
#define DefaultGcc2AxpOpt -O2 -mcpu=ev6
#define DefaultGcc2AxpOpt -O2 -mcpu=750
#define DefaultGcc2i386Opt -O2
That second line should be...
#define DefaultGcc2PpcOpt -O2 -mcpu=750
...I think.
-jwb
_
On Thu, 5 Apr 2001, Jeffrey W. Baker wrote:
> I just updated my tree and got this at the top of host.def:
>
> #define DefaultGcc2AxpOpt -O2 -mcpu=ev6
> #define DefaultGcc2AxpOpt -O2 -mcpu=750
> #define DefaultGcc2i386Opt -O2
>
> That second line should be...
>
> #define DefaultGcc2PpcOpt -O2 -mcp
On Thu, 5 Apr 2001, Jeffrey W. Baker wrote:
> On Thu, 5 Apr 2001, Jeffrey W. Baker wrote:
>
> > I just updated my tree and got this at the top of host.def:
> >
> > #define DefaultGcc2AxpOpt -O2 -mcpu=ev6
> > #define DefaultGcc2AxpOpt -O2 -mcpu=750
> > #define DefaultGcc2i386Opt -O2
> >
> > That s
"Jeffrey W. Baker" wrote:
>
> I just updated my tree and got this at the top of host.def:
>
> #define DefaultGcc2AxpOpt -O2 -mcpu=ev6
> #define DefaultGcc2AxpOpt -O2 -mcpu=750
> #define DefaultGcc2i386Opt -O2
>
> That second line should be...
>
> #define DefaultGcc2PpcOpt -O2 -mcpu=750
>
> ..
"Jeffrey W. Baker" wrote:
>
> On Thu, 5 Apr 2001, Jeffrey W. Baker wrote:
>
> > I just updated my tree and got this at the top of host.def:
> >
> > #define DefaultGcc2AxpOpt -O2 -mcpu=ev6
> > #define DefaultGcc2AxpOpt -O2 -mcpu=750
> > #define DefaultGcc2i386Opt -O2
> >
> > That second line shou
On Thu, Apr 05, 2001 at 06:04:11PM -0700, Jeffrey W. Baker wrote:
> I just updated my tree and got this at the top of host.def:
>
> #define DefaultGcc2AxpOpt -O2 -mcpu=ev6
> #define DefaultGcc2AxpOpt -O2 -mcpu=750
> #define DefaultGcc2i386Opt -O2
>
> That second line should be...
>
> #define De
Hello,
I have been trying to use Zephaniah E. Hull's Debian DRI packages
(versioned 2001.04.01-1) with my 8MB AGP G200 at 640x480. I've been
using the kernel module from DRI CVS.
Very soon (between 3 seconds and a minute) after starting any
XScreensaver GL hack, or QuakeForge-gl, that process se
On Thu, Apr 05, 2001 at 10:07:44PM -0700, Aaron Lehmann wrote:
> Very soon (between 3 seconds and a minute) after starting any
> XScreensaver GL hack, or QuakeForge-gl, that process segfaults in
> __driCreateScreen. All of the GL XScreensaver hacks work flawlessly with
> plain mesa. The backtraces
On Thu, Apr 05, 2001 at 10:24:13PM -0700, Aaron Lehmann wrote:
> FWIW everything seems to work OK in 24bpp mode.
Or not quite. I can still get quakeforge to segfault after a few
minutes, and tuxracer locks the machine up pretty quickly.
PGP signature
I've gotten into the habit of just doing
a 'make install', since the xserver is
pretty stable for my card. But what/where
are all the files (binaries/libs/modules)
that get overwritten when I do 'make install'
that I should be concerned about backing up?
Nick
--
Nobody will ever need more than
Nicholas Leippe wrote:
>
> I've gotten into the habit of just doing
> a 'make install', since the xserver is
> pretty stable for my card. But what/where
> are all the files (binaries/libs/modules)
> that get overwritten when I do 'make install'
> that I should be concerned about backing up?
When
35 matches
Mail list logo
