Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
2.6.29 and ATI drivers
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4  Next  
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware
View previous topic :: View next topic  
Author Message
milomak
Apprentice
Apprentice


Joined: 10 Apr 2008
Posts: 287

PostPosted: Fri May 22, 2009 7:06 pm    Post subject: Reply with quote

am i having this same problem when I have to set the driver to vesa and when i use startxfce4, the mouse and keyboard is dead?

even though i have settings in /etc/hal/fdi/policy/10-x11-... that work on other distros?

Code:

[I] x11-drivers/ati-drivers
     Available versions:  [M]8.27.10-r1!s 8.32.5!s!t (~)8.33.6-r1!s!t [M]8.35.5 [M](~)8.36.5 [M]8.39.4 [M]8.40.4 [M](~)8.452 [M]8.471.3 8.552-r2 (~)8.552-r3 (~)8.573-r1 (~)8.582 (~)8.593 {acpi debug doc kernel_linux multilib opengl qt3}
     Installed versions:  8.593(20:04:28 05/22/09)(kernel_linux -acpi -debug)
     Homepage:            http://www.ati.com
     Description:         Ati precompiled drivers for recent chipsets

[I] x11-drivers/xf86-video-ati
     Available versions:  6.6.3 6.8.0-r1 (~)6.9.0 (~)6.10.0 (~)6.11.0 6.12.1-r1 (~)6.12.2 {debug}
     Installed versions:  6.12.2(20:05:20 05/22/09)(-debug)
     Homepage:            http://xorg.freedesktop.org/
     Description:         ATI video driver
[U] x11-base/xorg-server
     Available versions:  1.3.0.0-r6 (~)1.4.2 (~)1.5.3-r4 1.5.3-r5 (~)1.5.3-r6 {M}(~)1.6.1.901 {M}(~)1.6.1.901-r1 {3dfx debug dmx dri hal input_devices_acecad input_devices_aiptek input_devices_calcomp input_devices_citron input_devices_digitaledge input_devices_dmc input_devices_dynapro input_devices_elo2300 input_devices_elographics input_devices_evdev input_devices_fpit input_devices_hyperpen input_devices_jamstudio input_devices_joystick input_devices_keyboard input_devices_magellan input_devices_microtouch input_devices_mouse input_devices_mutouch input_devices_palmax input_devices_penmount input_devices_spaceorb input_devices_summa input_devices_synaptics input_devices_tek4957 input_devices_tslib input_devices_ur98 input_devices_virtualbox input_devices_vmmouse input_devices_void input_devices_wacom ipv6 kdrive minimal nptl sdl tslib video_cards_apm video_cards_ark video_cards_ast video_cards_chips video_cards_cirrus video_cards_cyrix video_cards_dummy video_cards_epson video_cards_fbdev video_cards_fglrx video_cards_geode video_cards_glint video_cards_i128 video_cards_i740 video_cards_impact video_cards_imstt video_cards_intel video_cards_mach64 video_cards_mga video_cards_neomagic video_cards_newport video_cards_nsc video_cards_nv video_cards_nvidia video_cards_r128 video_cards_radeon video_cards_radeonhd video_cards_rendition video_cards_s3 video_cards_s3virge video_cards_savage video_cards_siliconmotion video_cards_sis video_cards_sisusb video_cards_sunbw2 video_cards_suncg14 video_cards_suncg3 video_cards_suncg6 video_cards_sunffb video_cards_sunleo video_cards_suntcx video_cards_tdfx video_cards_tga video_cards_trident video_cards_tseng video_cards_v4l video_cards_vermilion video_cards_vesa video_cards_vga video_cards_via video_cards_virtualbox video_cards_vmware video_cards_voodoo video_cards_xgi xorg}
     Installed versions:  1.6.1.901(02:07:43 05/22/09)(hal input_devices_evdev input_devices_keyboard input_devices_mouse ipv6 nptl video_cards_fglrx video_cards_radeon video_cards_radeonhd video_cards_v4l video_cards_vesa xorg -debug -dmx -input_devices_acecad -input_devices_aiptek -input_devices_calcomp -input_devices_citron -input_devices_digitaledge -input_devices_dmc -input_devices_dynapro -input_devices_elo2300 -input_devices_elographics -input_devices_fpit -input_devices_hyperpen -input_devices_jamstudio -input_devices_joystick -input_devices_magellan -input_devices_microtouch -input_devices_mutouch -input_devices_palmax -input_devices_penmount -input_devices_spaceorb -input_devices_summa -input_devices_synaptics -input_devices_tek4957 -input_devices_tslib -input_devices_ur98 -input_devices_virtualbox -input_devices_vmmouse -input_devices_void -input_devices_wacom -kdrive -minimal -sdl -tslib -video_cards_apm -video_cards_ark -video_cards_ast -video_cards_chips -video_cards_cirrus -video_cards_dummy -video_cards_epson -video_cards_fbdev -video_cards_geode -video_cards_glint -video_cards_i128 -video_cards_i740 -video_cards_impact -video_cards_imstt -video_cards_intel -video_cards_mach64 -video_cards_mga -video_cards_neomagic -video_cards_nv -video_cards_nvidia -video_cards_r128 -video_cards_rendition -video_cards_s3 -video_cards_s3virge -video_cards_savage -video_cards_siliconmotion -video_cards_sis -video_cards_sisusb -video_cards_sunbw2 -video_cards_suncg14 -video_cards_suncg3 -video_cards_suncg6 -video_cards_sunffb -video_cards_sunleo -video_cards_suntcx -video_cards_tdfx -video_cards_tga -video_cards_trident -video_cards_tseng -video_cards_vermilion -video_cards_via -video_cards_virtualbox -video_cards_vmware -video_cards_voodoo -video_cards_xgi)
     Homepage:            http://xorg.freedesktop.org/
     Description:         X.Org X servers
Back to top
View user's profile Send private message
milomak
Apprentice
Apprentice


Joined: 10 Apr 2008
Posts: 287

PostPosted: Fri May 29, 2009 6:21 pm    Post subject: Reply with quote

aaargh, I didn't have hald starting at boot :oops:
Back to top
View user's profile Send private message
eis
Apprentice
Apprentice


Joined: 05 Aug 2005
Posts: 161
Location: Stuttgart, Germany

PostPosted: Sun May 31, 2009 2:28 pm    Post subject: Reply with quote

Hi everyone!

This is an interesing thread and it seems to at least touch my problem as well. Here it goes:

I use the gentoo 2.6.27-r8 kernel, just as the stable portage tree offers it.
Ich compiled ati-drivers without errors.

When I set up the xdm (I tried both xdm and slim as the manager), I always get a welcome screen that freezes my whole system. I can only reset.

In my kernel I chose the following options, just as described in this gentoo-ati-howto.

---
Processor type and features --->
<*> MTRR (Memory Type Range Register) support
Device drivers --->
Graphics support --->
<M> /dev/agpgart (AGP Support) --->
(The agpgart option is not present on 64-bit kernels; just choose your chipset support.)
<M> Intel 440LX/BX/GX, I8xx and E7x05 support
(Enable your chipset instead of the above.)
<M> Direct Rendering Manager (XFree86 4.1.0 and higher DRI support) --->
<M> (Select your graphics card from the list)
---

I have a 32 bit dell laptop with an ati mobilitiy X300 graphics card.

Does anyone know what I am doing wrong?

greetings,

boris
Back to top
View user's profile Send private message
milomak
Apprentice
Apprentice


Joined: 10 Apr 2008
Posts: 287

PostPosted: Sun May 31, 2009 4:06 pm    Post subject: Reply with quote

how can i find out what package is trying to pull ati-drivers even though I do not have fglrx in VIDEOCARDS?
Back to top
View user's profile Send private message
ShinyDoofy
n00b
n00b


Joined: 22 Jul 2006
Posts: 73

PostPosted: Sun May 31, 2009 4:39 pm    Post subject: Reply with quote

Either use the -t option on emerge or try "equery d ati-drivers"
Back to top
View user's profile Send private message
eis
Apprentice
Apprentice


Joined: 05 Aug 2005
Posts: 161
Location: Stuttgart, Germany

PostPosted: Mon Jun 01, 2009 3:56 pm    Post subject: Reply with quote

hm..
still no solution for my problem. I never really thought about what kernel to use.
Usually I set all the options by hand, compile and install modules.
For updates I save the kernel configuration of the old kernel.

Is there a better way to go?

thanks in advance,

boris
Back to top
View user's profile Send private message
emuller
n00b
n00b


Joined: 28 Oct 2005
Posts: 23

PostPosted: Tue Jun 09, 2009 1:02 pm    Post subject: Reply with quote

ShinyDoofy wrote:
I did some more testing and it turned out that /etc/ati/amdpcsdb is a really critical file to have. When I delete it and start X, my screen dies and I can only REISUB (or log in via ssh and see what's going wrong). When I copy the old file back I had some time ago, everything works well (for kernel 2.6.28.4 and Catalyst 9.4).

Anyway, Catalyst 9.5 running for me for now with kernel 2.6.29.4 with this patch (although it's spamming dmesg with error messages).

/Also, for whatever reason, amdcccle doesn't segfault anymore when starting it.


For exactly this reason, I find it useful to put my fglrx config files under revision control:

cd /etc
hg init .
hg add /etc/X11/xorg.conf
hg add /etc/ati
hg commit

I then commit working amdpcsdb+xorg.conf configs ...
Back to top
View user's profile Send private message
Snake
Tux's lil' helper
Tux's lil' helper


Joined: 21 Jul 2005
Posts: 146

PostPosted: Tue Jun 09, 2009 9:31 pm    Post subject: Reply with quote

I planned to upgrade to 2.6.29 from 2.6.27 in following days, but after seeing that topic, I think I'll stick to good old 27 :wink: and wait for 2.6.30 to become stable and hope everything will be ok then.
Back to top
View user's profile Send private message
VoidMage
Watchman
Watchman


Joined: 14 Oct 2006
Posts: 6196

PostPosted: Tue Jun 09, 2009 11:12 pm    Post subject: Reply with quote

@Snake: and just what will that change ? -
it's ati-drivers that are broken, not kernel.
Back to top
View user's profile Send private message
adrs0061
n00b
n00b


Joined: 08 Jun 2009
Posts: 15

PostPosted: Wed Jun 10, 2009 3:45 am    Post subject: Reply with quote

Im not sure why most here are using the closed source driver and maybe you have valid reasons i dont know but from where i stand i can only say that the opensource drivers (have them running in all of my computers) are the way to go there stable and dont make any hassle when updating the kernel or the xorg-server, and from now on there the only drivers which continue to support R500 and below

and also with all the work which is currently done around ATI KMS and radeon driver gallium 3d support the opensource drivers are definitely the way to go

just wanted to say that because i think most of you would have less problems if you don't need 3D support at the moment with the opensource drivers

regards
Back to top
View user's profile Send private message
DaggyStyle
Watchman
Watchman


Joined: 22 Mar 2006
Posts: 5941

PostPosted: Wed Jun 10, 2009 4:03 am    Post subject: Reply with quote

adrs0061 wrote:
Im not sure why most here are using the closed source driver and maybe you have valid reasons i dont know but from where i stand i can only say that the opensource drivers (have them running in all of my computers) are the way to go there stable and dont make any hassle when updating the kernel or the xorg-server, and from now on there the only drivers which continue to support R500 and below

and also with all the work which is currently done around ATI KMS and radeon driver gallium 3d support the opensource drivers are definitely the way to go

just wanted to say that because i think most of you would have less problems if you don't need 3D support at the moment with the opensource drivers

regards


that is why.
_________________
Only two things are infinite, the universe and human stupidity and I'm not sure about the former - Albert Einstein
Back to top
View user's profile Send private message
i92guboj
Bodhisattva
Bodhisattva


Joined: 30 Nov 2004
Posts: 10315
Location: Córdoba (Spain)

PostPosted: Wed Jun 10, 2009 5:34 am    Post subject: Reply with quote

Besides that, both radeonhd and radeon fails miserably for a lot of people, just like fglrx.

Even if you don't need 3d, you might have the *luck* that only fglrx works for you. The situation of ATi cards is chaotic, as always.

I have a 2600 on one computer, and both radeon and radeonhd crash my server at startup, causing one of my monitors to go standby and similar things. Yeah, ATi is specially daunting on setups with more than one monitor.
Back to top
View user's profile Send private message
radio_flyer
Guru
Guru


Joined: 04 Nov 2004
Posts: 318
Location: Northern California

PostPosted: Wed Jun 10, 2009 6:41 am    Post subject: Reply with quote

3D 3D 3D

Not for Compiz. Not for gaming. For engineering. Serious engineering work requires OpenGL.

I've run the radeon driver in the past on a 9200. Great 3D support. Loved it. On my current HD2600XT, I'm running fglrx ( and stuck at 2.6.28 ) because 3D is not optional. I'm running an ATI card solely because I know open-source R600 3D support is coming, thanks to ATI and the X developers. The moment usable open source 3D shows up in either the radeon or radeonhd driver, fglrx is *so* off my system...
Back to top
View user's profile Send private message
emuller
n00b
n00b


Joined: 28 Oct 2005
Posts: 23

PostPosted: Wed Jun 10, 2009 7:32 am    Post subject: Reply with quote

I am using the AMD Stream SDK, which requires fglrx

For R700 based hardware (4850,4870,4870x2), the fglrx driver can be made to work. Stick with Xorg 1.5.3 or 1.5.2 for the time being (no >=1.6) and <=2.6.28 kernel.

Funny behaviour is usually due to /etc/ati/amdpcsdb

Generate your xorg.conf like so:

rm /etc/X11/xorg.conf
touch /etc/X11/xorg.conf

aticonfig --initial --adapter=all -f

But this is not a "fresh" config as the amdpcsdb remains. Deleting that can sometimes improve the situation.
Best thing is to keep all this stuff under revision control so one doesn't have to reinstall to get back to a working config ... see above post.

My trouble is I want to get more than two 4870x2 cards working with fglrx for stream computing. With that, I have not yet succeeded.
Back to top
View user's profile Send private message
adrs0061
n00b
n00b


Joined: 08 Jun 2009
Posts: 15

PostPosted: Wed Jun 10, 2009 11:18 pm    Post subject: Reply with quote

in my experience i had a lot less problems with the opensource radeon driver, at least with the released versions, i dont use the radeonhd much, but i can't speak for anything newer then a RS600 and i don't use any 3D at, and the only experience with two monitor support is the RS600 which works with the radeon driver like a charm

excuse my shortsightedness on some of these issues :S
Back to top
View user's profile Send private message
h2sammo
Veteran
Veteran


Joined: 11 Mar 2009
Posts: 1025
Location: Michigan

PostPosted: Thu Jun 11, 2009 7:28 pm    Post subject: Reply with quote

so, 2.6.30 was released.

1. does 2.6.29 work with ATI w/out additional tinkering?

2. does 2.6.30 support ATI video cards?
Back to top
View user's profile Send private message
adrs0061
n00b
n00b


Joined: 08 Jun 2009
Posts: 15

PostPosted: Sun Jun 14, 2009 2:49 am    Post subject: Reply with quote

2.6.30 does support radeon cards up to the R700 with 2D support (opensource), i dont think theres any fglrx support yet

im not sure what u mean with point 1. but theres no official 2.6.29 support within the fglrx driver

regards
Back to top
View user's profile Send private message
theRealMorpheu5
l33t
l33t


Joined: 26 Feb 2004
Posts: 841
Location: Vicenza, IT

PostPosted: Sat Jun 20, 2009 9:39 am    Post subject: Reply with quote

According to http://linux.com/community/blogs/ATI-Catalyst-fglrx-and-Kernels-2.6.30-2.6.29.html there seems to be a hope.

Is this happening anytime soon? Is anyone having random lockups (kernel panic with flashing caps-lock light and no more details) when using various xvideo/opengl applications? For example: I cannot play movies with Mplayer or work with Blender for long time since they, sooner or later, make the kernel panic.

EDIT: I'm using ati-drivers-8.593 and all the stable (not keyworded) Xorg things.

EDIT 2: Forgot to mention that I'm happily running kde-4.2.4-r2 with compositing enabled, and glxgears, as well as fgl_glxgears are apparently not causing any crash.
Back to top
View user's profile Send private message
DevSolar
Tux's lil' helper
Tux's lil' helper


Joined: 07 Sep 2004
Posts: 82

PostPosted: Sun Jun 21, 2009 8:38 am    Post subject: Reply with quote

VoidMage wrote:
it's ati-drivers that are broken, not kernel.


It's the kernel policy of changing API / ABI between versions on a whim that's broken, and breaking many things for many people for years now.
Back to top
View user's profile Send private message
nego
n00b
n00b


Joined: 28 Jun 2005
Posts: 6

PostPosted: Sun Jun 21, 2009 9:41 am    Post subject: Reply with quote

I have a ATI Sapphire 3850hd and the open source ati radeon drivers work flawlessly with the automatically generated xorg config. 2d and 3d acceleration work about as well as I've experienced with fglrx drivers. I only tried them after I could not manage to get fglrx working with the latest kernel.
Back to top
View user's profile Send private message
sylvain_
n00b
n00b


Joined: 13 Aug 2007
Posts: 56

PostPosted: Sun Jun 21, 2009 11:38 am    Post subject: rseuib Reply with quote

ShinyDoofy wrote:

and Alt+SysRq+REISUB still work great, but that's about it.


that's RSEUIB (yeah that dosent fell right)
_________________
hunhun ça vaux mieux que pan tu l'aura
Back to top
View user's profile Send private message
DaggyStyle
Watchman
Watchman


Joined: 22 Mar 2006
Posts: 5941

PostPosted: Sun Jun 21, 2009 2:03 pm    Post subject: Reply with quote

nego wrote:
I have a ATI Sapphire 3850hd and the open source ati radeon drivers work flawlessly with the automatically generated xorg config. 2d and 3d acceleration work about as well as I've experienced with fglrx drivers. I only tried them after I could not manage to get fglrx working with the latest kernel.


well both 2d and 3d are software render not hardware render.
there is not software acceleration from these chips.
_________________
Only two things are infinite, the universe and human stupidity and I'm not sure about the former - Albert Einstein
Back to top
View user's profile Send private message
energyman76b
Advocate
Advocate


Joined: 26 Mar 2003
Posts: 2048
Location: Germany

PostPosted: Tue Jun 23, 2009 10:52 pm    Post subject: Reply with quote

with my HD3870, 2.6.29.X, catalyst 9.5&9.6

I need this patch:
diff -Nparu build_mod/firegl_public.c fglrx-8.573-new/common/lib/modules/fglrx/build_mod/firegl_public.c
--- build_mod/firegl_public.c 2009-01-23 20:00:26.000000000 -0200
+++ fglrx-8.573-new/common/lib/modules/fglrx/build_mod/firegl_public.c 2009-02-13 15:25:00.000000000 -0200
@@ -1460,7 +1460,11 @@ KCL_TYPE_Pid ATI_API_CALL KCL_GetTgid(vo
*/
KCL_TYPE_Uid ATI_API_CALL KCL_GetEffectiveUid(void)
{
+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,28)
+ return current->cred->euid;
+#else
return current->euid;
+#endif
}

/** /brief Delay execution for the specified number of microseconds
@@ -1832,15 +1836,30 @@ int ATI_API_CALL KCL_PosixSecurityCapChe
*/
void ATI_API_CALL KCL_PosixSecurityCapSetIPCLock(unsigned int lock)
{
+
+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,28)
+ struct cred *new = prepare_creds();
+ if (!new) {
+ printk(KERN_ERR "fglrx: could not allocate memory\n");
+ return;
+ }
+#else
+#define new current
+#endif
if (lock == 0 )
{
- cap_lower(current->cap_effective, CAP_IPC_LOCK);
+ cap_lower(new->cap_effective, CAP_IPC_LOCK);
}
else
{
- cap_raise(current->cap_effective, CAP_IPC_LOCK);
+ cap_raise(new->cap_effective, CAP_IPC_LOCK);
}
- return;
+
+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,28)
+ commit_creds(new);
+#else
+#undef new
+#endif
}

/** \brief Get number of available RAM pages
diff -Nparu build_mod/firegl_public.h fglrx-8.573-new/common/lib/modules/fglrx/build_mod/firegl_public.h
--- build_mod/firegl_public.h 2009-01-23 20:00:26.000000000 -0200
+++ fglrx-8.573-new/common/lib/modules/fglrx/build_mod/firegl_public.h 2009-02-16 14:02:50.000000000 -0300
@@ -18,6 +18,7 @@
#define _FIREGL_PUBLIC_H_

#include <stdarg.h>
+#include <asm/pgtable.h>
#include "kcl_pci.h"
#include "kcl_io.h"

@@ -590,6 +591,11 @@ extern unsigned long KCL_SYSINFO_
#define cpu_has_pge test_bit(X86_FEATURE_PGE, &boot_cpu_data.x86_capability)
#endif

+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,28)
+#undef pgprot_writecombine
+#undef pgprot_noncached
+#endif
+
#ifndef pgprot_writecombine
#define pgprot_writecombine(prot) __pgprot((pgprot_val(prot) & ~(_PAGE_PCD)) | _PAGE_PWT)
#endif
@@ -598,6 +604,7 @@ extern unsigned long KCL_SYSINFO_
#define pgprot_noncached(prot) __pgprot(pgprot_val(prot) | _PAGE_PCD | _PAGE_PWT)
#endif

+
#endif //FIREGL_USWC_SUPPORT


diff -Nparu build_mod/kcl_acpi.c fglrx-8.573-new/common/lib/modules/fglrx/build_mod/kcl_acpi.c
--- build_mod/kcl_acpi.c 2009-01-23 20:00:26.000000000 -0200
+++ fglrx-8.573-new/common/lib/modules/fglrx/build_mod/kcl_acpi.c 2009-02-13 15:25:00.000000000 -0200
@@ -18,6 +18,12 @@
#include <linux/autoconf.h>
#include <linux/acpi.h>

+#if LINUX_VERSION_CODE > KERNEL_VERSION(2,6,28)
+#include <../drivers/acpi/acpica/acconfig.h>
+#include <../drivers/acpi/acpica/aclocal.h>
+#include <../drivers/acpi/acpica/acobject.h>
+#endif
+
#include "kcl_config.h"
#include "kcl_type.h"
#include "kcl_acpi.h"
#include "kcl_acpi.h"

after that 3d&co works but dmesg is spammed with this:
[14734.504469] [fglrx:firegl_find_any_map] *ERROR* Invalid map handle!<3>[fglrx:drm_vm_open] *ERROR* map not found -> inconsistent kernel data!!! vma_start:0x7fde13660000,handle:0xfdff0000
[14734.506516] [fglrx:firegl_find_any_map] *ERROR* Invalid map handle!<3>[fglrx:drm_vm_open] *ERROR* map not found -> inconsistent kernel data!!! vma_start:0x7fde12dbf000,handle:0xd11ac000
[14734.510144] [fglrx:firegl_find_any_map] *ERROR* Invalid map handle!<3>[fglrx:drm_vm_close] *ERROR* map not found -> inconsistent kernel data!!! vma_start:0x7fde12dbf000,handle:0xd11ac000
[14735.026129] [fglrx:firegl_find_any_map] *ERROR* Invalid map handle!<3>[fglrx:drm_vm_open] *ERROR* map not found -> inconsistent kernel data!!! vma_start:0x7fddf7d4d000,handle:0xd11f4000
_________________
Study finds stunning lack of racial, gender, and economic diversity among middle-class white males

I identify as a dirty penismensch.
Back to top
View user's profile Send private message
i92guboj
Bodhisattva
Bodhisattva


Joined: 30 Nov 2004
Posts: 10315
Location: Córdoba (Spain)

PostPosted: Tue Jun 23, 2009 11:04 pm    Post subject: Reply with quote

DevSolar wrote:
VoidMage wrote:
it's ati-drivers that are broken, not kernel.


It's the kernel policy of changing API / ABI between versions on a whim that's broken, and breaking many things for many people for years now.


That's completely irrelevant. The kernel development can't remain stagnated due to the release cycles of external projects (and much less when they're closed source projects and the kernel developers can't do anything about that). So your solution would be holding back the whole community because ATI -whom don't even belong to this opensource community- can't keep the pace? I hope that on a second guess you don't really think that way.

When the api breaks is for a good reason, because *something* is happening in that regard in the kernel, like kms and a lot of other goodies that are coming into scene right now. It's not like the developers enjoy breaking the api to annoy the external developers...

Blame ATi or use an older kernel. They are well known because they always lag behind the kernel and xorg. nVidia can keep the pace and so should ATi, it's just that their driver is such a blob that not even them can maintain it within a reasonable time frame.
Back to top
View user's profile Send private message
DevSolar
Tux's lil' helper
Tux's lil' helper


Joined: 07 Sep 2004
Posts: 82

PostPosted: Wed Jun 24, 2009 2:30 pm    Post subject: Reply with quote

i92guboj wrote:
DevSolar wrote:
VoidMage wrote:
it's ati-drivers that are broken, not kernel.


It's the kernel policy of changing API / ABI between versions on a whim that's broken, and breaking many things for many people for years now.


That's completely irrelevant.


I actually believe that to be very relevant, nay, at the core of what deficencies Linux still has before it can be a truly great OS.

Quote:
The kernel development can't remain stagnated due to the release cycles of external projects (and much less when they're closed source projects and the kernel developers can't do anything about that).


Stable / backward-compatible / legacy-supported APIs != stagnated development.

Forward compatible / extensible API design, clean upgrade paths and testing phases including third-party vendors == good software engineering.

(This isn't just targeting the LKM's, but the Gentoo maintainers, too. The latest stable ati-drivers doesn't compile with the latest stable gentoo-sources. So what's "stable" here?)

Quote:
So your solution would be holding back the whole community because ATI -whom don't even belong to this opensource community- can't keep the pace?


First, it doesn't matter squat if someone belongs to "the opensource community" or not. An operating system is about computing productivity, not about philosophy or communities.

Second, you are talking about "keeping the pace". Was there advance warning in due time that a relevant API would change, including comprehensive information about the new API?

Since they did away with the "2.<even> for stable, 2.<uneven> for experimental" scheme, things have gone from bad to worse IMHO.

Quote:
When the api breaks is for a good reason, because *something* is happening in that regard in the kernel, like kms and a lot of other goodies that are coming into scene right now.


I (and I assume many others here) have, before they bought their hardware, checked that it actually works with Linux. Now it works better for some, but not at all for others.

Quote:
It's not like the developers enjoy breaking the api to annoy the external developers...


Even if those external developers "don't belong to this community"? I'm not so sure about that.

Quote:
Blame ATi or use an older kernel.


Will the LKM's provide security patches for my older kernel for, uh, what's usual in the business, two years or so, minimum? Will security patches by other package maintainers be compatible with my older kernel for that time period?

No, they won't.

Thanks to the fact that the LKM's consider a stable API "nonsense" (as documented in /usr/src/linux/Documentation), Linux is, and probably always will be, the OS that breaks every now and again because the kernel gets out of sync with the "supporting cast". XFree86, wireless, now it's the graphics driver. I'm curious as to what it will be next time.

Quote:
They are well known because they always lag behind the kernel and xorg. nVidia can keep the pace and so should ATi...


Well, back when I last thought about which hardware to buy, it was the other way 'round - if you wanted 3D hardware, ATI was the ticket and nVidia sucked. Such things change, the hardware on my desk doesn't, at least not without investing in new shiny.

I don't say ATI is free of blame. but the LKM's have to accept their share of it.
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware All times are GMT
Goto page Previous  1, 2, 3, 4  Next
Page 2 of 4

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum