SiS630 and 5591/5592 AGP

SiS630 and 5591/5592 AGP

[Gavin Baker]

My new laptop has this combination and the sis framebuffer driver
mangles the display. The sis X driver produces a nice lavalamp style
pattern that fades to white, the kernel stays alive but the machine
needs a reboot to fix the display in both cases. 

Im guessing the lack of 5591/5592 AGP support is the problem, and I 
was just wondering if anyone is working on this or should i go bug SiS?

Everything else is 100%.

00:00.0 Host bridge: Silicon Integrated Systems [SiS] 630 Host (rev 31)
00:00.1 IDE interface: Silicon Integrated Systems [SiS] 5513 [IDE] (rev
d0)
00:01.0 ISA bridge: Silicon Integrated Systems [SiS] 85C503/5513
00:01.1 Ethernet controller: Silicon Integrated Systems [SiS] SiS900
10/100 Ethernet (rev 82)
00:01.2 USB Controller: Silicon Integrated Systems [SiS] 7001 (rev 07)
00:01.3 USB Controller: Silicon Integrated Systems [SiS] 7001 (rev 07)
00:01.4 Multimedia audio controller: Silicon Integrated Systems [SiS]
SiS PCI Audio Accelerator (rev 02)
00:01.6 Modem: Silicon Integrated Systems [SiS]: Unknown device 7013
(rev a0)
00:02.0 PCI bridge: Silicon Integrated Systems [SiS] 5591/5592 AGP
00:03.0 CardBus bridge: O2 Micro, Inc. OZ6812 Cardbus Controller (rev
05)
01:00.0 VGA compatible controller: Silicon Integrated Systems [SiS]
SiS630 GUI Accelerator+3D (rev 31)

Cheers,
Gavin Baker

Re: SiS630 and 5591/5592 AGP

[Stuart Young]

At 06:59 PM 11/11/01 +0000, Gavin Baker wrote:
>My new laptop has this combination and the sis framebuffer driver
>mangles the display. The sis X driver produces a nice lavalamp style
>pattern that fades to white, the kernel stays alive but the machine
>needs a reboot to fix the display in both cases.
>
>Im guessing the lack of 5591/5592 AGP support is the problem, and I
>was just wondering if anyone is working on this or should i go bug SiS?

You've got the same problem I have. There is something weird with the way 
some laptops (and LCD based desktops) handle this display, and the only way 
I've gotten around it is to use the VesaFB driver (which is heaps slower 
than the accelerated driver), and then either use the FB driver for X 
(which is buggy IMHO, crashes if it gets too much to do buffered up), or 
disable the mode changes and use the SiS accelerated X driver.

Alan has new drivers in his ac tree. but I've tried them, and no luck. Give 
them a shot and see how you go. You might be lucky.

Out of interest, what model/manufacturer is the notebook? The machines I'm 
having problems with are Clevo LP200S's, which is an upright LCD machine, 
with the h/drive and power supply down in the stand/foot.

Once we get everything working with Linux, they'll suit the application we 
have for them rather nicely. *sigh*

Q: Slightly related, have you gotten the sound to work? Driver loads, I get 
"Unknown Codec" and an empty codec value. Reloading the driver makes no 
difference (even numerous times). Wondering if your system is the same...


AMC Enterprises P/L    - Stuart Young
First Floor            - Network and Systems Admin
3 Chesterville Rd      - sgy@amc.com.au
Cheltenham Vic 3192    - Ph:  (03) 9584-2700
http://www.amc.com.au/ - Fax: (03) 9584-2755

Re: SiS630 and 5591/5592 AGP

[Marco Colombo]

On Mon, 12 Nov 2001, Stuart Young wrote:

> At 06:59 PM 11/11/01 +0000, Gavin Baker wrote:
> >My new laptop has this combination and the sis framebuffer driver
> >mangles the display. The sis X driver produces a nice lavalamp style
> >pattern that fades to white, the kernel stays alive but the machine
> >needs a reboot to fix the display in both cases.
> >
> >Im guessing the lack of 5591/5592 AGP support is the problem, and I
> >was just wondering if anyone is working on this or should i go bug SiS?
> 
> You've got the same problem I have. There is something weird with the way 
> some laptops (and LCD based desktops) handle this display, and the only way 
> I've gotten around it is to use the VesaFB driver (which is heaps slower 
> than the accelerated driver), and then either use the FB driver for X 
> (which is buggy IMHO, crashes if it gets too much to do buffered up), or 
> disable the mode changes and use the SiS accelerated X driver.
> 

00:00.0 Host bridge: Silicon Integrated Systems [SiS] 630 Host (rev 11)
00:00.1 IDE interface: Silicon Integrated Systems [SiS] 5513 [IDE] (rev d0)
00:01.0 ISA bridge: Silicon Integrated Systems [SiS] 85C503/5513
00:01.1 Ethernet controller: Silicon Integrated Systems [SiS] SiS900 10/100 Ethernet (rev 80)
00:01.2 USB Controller: Silicon Integrated Systems [SiS] 7001 (rev 07)
00:01.4 Multimedia audio controller: Silicon Integrated Systems [SiS] SiS PCI Audio Accelerator (rev 01)
00:01.6 Modem: Silicon Integrated Systems [SiS]: Unknown device 7013 (rev a0)
00:02.0 PCI bridge: Silicon Integrated Systems [SiS] 5591/5592 AGP
00:08.0 CardBus bridge: Texas Instruments PCI1225 (rev 01)
00:08.1 CardBus bridge: Texas Instruments PCI1225 (rev 01)
01:00.0 VGA compatible controller: Silicon Integrated Systems [SiS] SiS630 GUI Accelerator+3D (rev 11)

The X server is Free86-4.0.3 (RH 7.1), Driver "fbdev", vga=0x317 as boot
param (VESA fb driver):

vesafb: framebuffer at 0xf0000000, mapped to 0xcc000000, size 8192k
vesafb: mode is 1024x768x16, linelength=2048, pages=4
vesafb: protected mode interface info at ca5b:0004
vesafb: scrolling: redraw
vesafb: directcolor: size=0:5:6:5, shift=0:11:5:0

I'm not able to run X at 24/32 bits depth with fbdev, and it works only at
8 bit without fbdev (native sis driver). I haven't tried the DRM module.

BTW, does anybody know which value I have to pass to vga= for
1024x768x24?

I've been using the fbdev for a while, no crashes so far.
Textmode (Alt-Fx) is restored correctly, too.

> Alan has new drivers in his ac tree. but I've tried them, and no luck. Give 
> them a shot and see how you go. You might be lucky.

I've been switch from vanilla 2.4.9 to 2.4.10-ac10, to 2.4.9-12 (RH update)
without any problem. It hangs (at PCMCIA initialization I believe) after
a warm reboot from Win2000, but it happens only by mistake, as I know I
should be cold rebooting it when switching OS.

> Out of interest, what model/manufacturer is the notebook? The machines I'm 
> having problems with are Clevo LP200S's, which is an upright LCD machine, 
> with the h/drive and power supply down in the stand/foot.

Mitac 7521.

> Once we get everything working with Linux, they'll suit the application we 
> have for them rather nicely. *sigh*

It does for me. 1024x768x16 looks fine at 14.1", and I have no real need
for 24 bits.

> Q: Slightly related, have you gotten the sound to work? Driver loads, I get 
> "Unknown Codec" and an empty codec value. Reloading the driver makes no 
> difference (even numerous times). Wondering if your system is the same...

Yes, all fine here with standard RH config.

.TM.
-- 
      ____/  ____/   /
     /      /       /			Marco Colombo
    ___/  ___  /   /		      Technical Manager
   /          /   /			 ESI s.r.l.
 _____/ _____/  _/		       Colombo@ESI.it

Re: SiS630 and 5591/5592 AGP

[Gavin Baker]

> >My new laptop has this combination and the sis framebuffer driver
> >mangles the display. The sis X driver produces a nice lavalamp style
> >pattern that fades to white, the kernel stays alive but the machine
> >needs a reboot to fix the display in both cases.
> >
> >Im guessing the lack of 5591/5592 AGP support is the problem, and I
> >was just wondering if anyone is working on this or should i go bug SiS?
> 
> You've got the same problem I have. There is something weird with the way 
> some laptops (and LCD based desktops) handle this display, and the only way 
> I've gotten around it is to use the VesaFB driver (which is heaps slower 
> than the accelerated driver), and then either use the FB driver for X 
> (which is buggy IMHO, crashes if it gets too much to do buffered up), or 
> disable the mode changes and use the SiS accelerated X driver.

Im currently using vesa for X, I would like to get some kind of acceleration
to play DVD's etc. I've tried the sis_drv in X 4.1 with the lavalamp
style effect i mentioned, the sis driver in X3.x gets the screen offset
totally wrong, ie the center of the screen is on the far right of the
physical screen, but this is a problem i can direct to the X lists.
 
> Alan has new drivers in his ac tree. but I've tried them, and no luck. Give 
> them a shot and see how you go. You might be lucky.

This is on 2.4.13-ac8, sisfb also messes up the screen offset,
CONFIG_AGP_SIS is turned off as Configure.help mentions it doesn't
support the 5591/5592.

> Out of interest, what model/manufacturer is the notebook? The machines I'm 
> having problems with are Clevo LP200S's, which is an upright LCD machine, 
> with the h/drive and power supply down in the stand/foot.

This is an "Advent 5490", which i think is Pc-World(tm)'s own brand name
they've applied to an oem laptop made by someone else, but i could have
this the wrong way around/totally wrong. It's a 900mhz celeron (coppermine), 
SiS everything.

Im sure other laptops with different badges on them will have the same
internals.

> Once we get everything working with Linux, they'll suit the application we 
> have for them rather nicely. *sigh*

I've mailed SiS asking them if they have any support planned, no reply
as yet, heres to hoping...

> Q: Slightly related, have you gotten the sound to work? Driver loads, I get 
> "Unknown Codec" and an empty codec value. Reloading the driver makes no 
> difference (even numerous times). Wondering if your system is the same...

Everything else (except the software modem, which i have no use for)
worked straight out of the tin with RedHat 7.2's vanilla kernel. Im
current using Alans 2.4.13-ac8:

/sbin/modprobe trident

Trident 4DWave/SiS 7018/ALi 5451,Tvia CyberPro 5050 PCI Audio, version
0.14.9d,
23:06:41 Nov 11 2001
PCI: Found IRQ 11 for device 00:01.4
PCI: Sharing IRQ 11 with 00:03.0
trident: SiS 7018 PCI Audio found at IO 0xd400, IRQ 11
ac97_codec: AC97 Audio codec, id: 0x414c:0x4710 (ALC200/200P)
ac97_codec: AC97 Modem codec, id: 0x5349:0x4c22 (Silicon Laboratory
Si3036)

It works perfectly, as does the NIC.

Cheers, Gavin Baker

Re: CONFIG_FB_SIS and >1 display Was: SiS630 and 5591/5592 AGP

[Gavin Baker]

Ok, the FB_SIS driver _does_ work on this machine, but sends its output to
the external monitor connector (I see all but the leftmost characters,
which the monitor cannot fix), but sends garbage to the builtin screen.

Any simple way to fix this, or does it fall under the "SiS wont tell us
how to fix it" catagory? As noted, this seems to be the problem X has
with it too.

Cheers, Gavin Baker

Re: SiS630 and 5591/5592 AGP

[Stuart Young]

At 04:36 PM 12/11/01 +0000, Gavin Baker wrote:
>Im currently using vesa for X, I would like to get some kind of 
>acceleration to play DVD's etc. I've tried the sis_drv in X 4.1 with the 
>lavalamp style effect i mentioned, the sis driver in X3.x gets the screen 
>offset totally wrong, ie the center of the screen is on the far right of 
>the physical screen, but this is a problem i can direct to the X lists.

You'll find a post from me there back in October, and a patch to X (and a 
binary module) linked to. The files are still there, and contains details 
on how to try the work-around.

> > Alan has new drivers in his ac tree. but I've tried them, and no luck.
> > Give them a shot and see how you go. You might be lucky.
>
>This is on 2.4.13-ac8, sisfb also messes up the screen offset,
>CONFIG_AGP_SIS is turned off as Configure.help mentions it doesn't
>support the 5591/5592.



>This is an "Advent 5490", which i think is Pc-World(tm)'s own brand name 
>they've applied to an oem laptop made by someone else, but i could have 
>this the wrong way around/totally wrong. It's a 900mhz 
>celeron  (coppermine), SiS everything.

Probably a Clevo, they make a lot of notebooks. Then again, you never know.

>Im sure other laptops with different badges on them will have the same
>internals.

Gotta love reference designs.

>Everything else (except the software modem, which i have no use for)
>worked straight out of the tin with RedHat 7.2's vanilla kernel. Im
>current using Alans 2.4.13-ac8:

The sisfb driver changes appeared in 2.4.13-ac4, so I'd guess they'd be in 
2.4.13-ac8. Unfortunately (afaik) Alan hasn't gotten much in the way of 
e-mail response from SiS on his questions. The driver made an appearance 
early in the 2.4 series, and was 'reverted' back to the old one as SiS 
didn't respond to questions about the code, which afaik seemed "inconsistent".

Can you e-mail me (may as well be off-list) the dump of 'lspci -bv' please? 
I'm interested in checking up on the revision differences between our 
chipsets. Should give me something to start with on the audio front, and 
hopefully on the video front I can get the same info from people without 
problems, so I can figure out what revisions have this problem. Any others 
who either have this video problem, or don't have it (SiS630 chipsets only 
at the moment, and I know a few people are keeping an eye out for topics 
with SiS630 in them like I am), feel free to e-mail me your output from 
'lspci -bv' as well. The more info I get, the more likely we can narrow 
this down.


AMC Enterprises P/L    - Stuart Young
First Floor            - Network and Systems Admin
3 Chesterville Rd      - sgy@amc.com.au
Cheltenham Vic 3192    - Ph:  (03) 9584-2700
http://www.amc.com.au/ - Fax: (03) 9584-2755

Re: SiS630 and 5591/5592 AGP

[Gavin Baker]

On Tue, Nov 13, 2001 at 11:15:11AM +1100, Stuart Young wrote:

> At 04:36 PM 12/11/01 +0000, Gavin Baker wrote:
> >Im currently using vesa for X, I would like to get some kind of 
> >acceleration to play DVD's etc. I've tried the sis_drv in X 4.1 with the 
> >lavalamp style effect i mentioned, the sis driver in X3.x gets the screen 
> >offset totally wrong, ie the center of the screen is on the far right of 
> >the physical screen, but this is a problem i can direct to the X lists.
> 
> You'll find a post from me there back in October, and a patch to X (and a 
> binary module) linked to. The files are still there, and contains details 
> on how to try the work-around.

Im using your patched sis_drv.o for X now. The performance increase is
noticable, but the Xv display still can't keep up with any highbandwidth video
(even 352x288 mpeg), so my dream of DVD playback is still far off but a lot
closer. ;)

> >Im sure other laptops with different badges on them will have the same
> >internals.
> 
> Gotta love reference designs.

Sure, when they are well supported reference designs :)

> >Everything else (except the software modem, which i have no use for)
> >worked straight out of the tin with RedHat 7.2's vanilla kernel. Im
> >current using Alans 2.4.13-ac8:
> 
> The sisfb driver changes appeared in 2.4.13-ac4, so I'd guess they'd be in 
> 2.4.13-ac8. Unfortunately (afaik) Alan hasn't gotten much in the way of 
> e-mail response from SiS on his questions. The driver made an appearance 
> early in the 2.4 series, and was 'reverted' back to the old one as SiS 
> didn't respond to questions about the code, which afaik seemed "inconsistent".

The SiS Framebuffer has the problem thats been mentioned where it
does'nt talk to the CRT/LCD controller correctly. It displays to the 
CRT _almost_ perfectly, but sends garbage to the LCD, i guess its the same 
for you. 

I will never understand the logic behind a company _not_ helping produce
drivers for their hardware under our favourite OS. The only reason i can
even think of is if their hardware needs huge ugly hacks that binary
drivers hide.
 
> Can you e-mail me (may as well be off-list) the dump of 'lspci -bv' please? 

done.

Cheers, Gavin Baker

Re: SiS630 and 5591/5592 AGP

[Alan Cox]

> Im using your patched sis_drv.o for X now. The performance increase is
> noticable, but the Xv display still can't keep up with any highbandwidth video
> (even 352x288 mpeg), so my dream of DVD playback is still far off but a lot
> closer. ;)

For the SiS 6326 you want mtrr's set up to make it write combining and
ideally the new Xvmc X11 support not just Xv