AMD Catalyst 12.6 Beta improvements?

[hobbes]

AMD took down 12.6 beta from the website.

Is there anyone here that could provide us a link to it?

Thanks

[Massa]

not sure if this is debian specific problem, because archlinux maintainer of catalyst has reported success with 3.4 Kernel as well as Xserver 1.12 .. but it looks like he is using a hack to make it compatible with kernel 3.4, not sure again because i have seen users within Phoronix forums reporting Ubuntu 12.04 with 3.4 kernel working with catalyst 12.6

https://bbs.archlinux.org/viewtopic.php?id=57084&p=159

Hi xgt001,

It's working for me under Fedora 17 64-bit with Xorg 1.12 and kernel 3.3.7-1.fc17.x86_64.

It's also quite snappier on my AMD E-350, very happy with this release so far.

[xgt001]

3.4 and xserver 1.12 working on arch

Hey guys,

we (Debian) just wanted to upload the fixed xserver 1.12 driver to Debian, but yeah the x 1.12 support is....:

Backtrace:
0: /usr/bin/X (xorg_backtrace+0x34) [0x7fbbe05283c4]
1: /usr/bin/X (0x7fbbe03a0000+0x18bf99) [0x7fbbe052bf99]
2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7fbbdf6c8000+0xf030)
[0x7fbbdf6d7030]
3: /usr/lib/x86_64-linux-gnu/libpciaccess.so.0 (0x7fbbdf8e4000+0x5394)
[0x7fbbdf8e9394]
4: /usr/lib/xorg/modules/drivers/fglrx_drv.so (0x7fbbdb7bb000+0x8032c9)
[0x7fbbdbfbe2c9]
5: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(amd_xs112_int10_xf86ExtendedInitInt10+0xdd) [0x7fbbdbfbe05d]
6: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(amd_xs112_int10_xf86InitInt10+0xb) [0x7fbbdbfbd8fb]
7: /usr/lib/xorg/modules/drivers/fglrx_drv.so (xf86InitInt10+0xd)
[0x7fbbdbb0397d]
8: /usr/lib/xorg/modules/drivers/fglrx_drv.so (0x7fbbdb7bb000+0x48405c)
[0x7fbbdbc3f05c]
9: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(xdl_xs112_PreInitAdapter+0x274) [0x7fbbdbc3ec34]
10: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(xdl_xs112_atiddxPreInit+0x12ad) [0x7fbbdbc4031d]
11: /usr/bin/X (InitOutput+0x83c) [0x7fbbe043287c]
12: /usr/bin/X (0x7fbbe03a0000+0x4181d) [0x7fbbe03e181d]
13: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xfd)
[0x7fbbde3f1ead]
14: /usr/bin/X (0x7fbbe03a0000+0x41cdd) [0x7fbbe03e1cdd]
Segmentation fault at address 0xffffffffe18c7332

... simply not working at all (on amd64 only maybe?)..............

not sure if this is debian specific problem, because archlinux maintainer of catalyst has reported success with 3.4 Kernel as well as Xserver 1.12 .. but it looks like he is using a hack to make it compatible with kernel 3.4, not sure again because i have seen users within Phoronix forums reporting Ubuntu 12.04 with 3.4 kernel working with catalyst 12.6

https://bbs.archlinux.org/viewtopic.php?id=57084&p=159

[the-me]

XServer 1.12 is still broken

Hey guys,

we (Debian) just wanted to upload the fixed xserver 1.12 driver to Debian, but yeah the x 1.12 support is....:

Backtrace:
0: /usr/bin/X (xorg_backtrace+0x34) [0x7fbbe05283c4]
1: /usr/bin/X (0x7fbbe03a0000+0x18bf99) [0x7fbbe052bf99]
2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7fbbdf6c8000+0xf030)
[0x7fbbdf6d7030]
3: /usr/lib/x86_64-linux-gnu/libpciaccess.so.0 (0x7fbbdf8e4000+0x5394)
[0x7fbbdf8e9394]
4: /usr/lib/xorg/modules/drivers/fglrx_drv.so (0x7fbbdb7bb000+0x8032c9)
[0x7fbbdbfbe2c9]
5: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(amd_xs112_int10_xf86ExtendedInitInt10+0xdd) [0x7fbbdbfbe05d]
6: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(amd_xs112_int10_xf86InitInt10+0xb) [0x7fbbdbfbd8fb]
7: /usr/lib/xorg/modules/drivers/fglrx_drv.so (xf86InitInt10+0xd)
[0x7fbbdbb0397d]
8: /usr/lib/xorg/modules/drivers/fglrx_drv.so (0x7fbbdb7bb000+0x48405c)
[0x7fbbdbc3f05c]
9: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(xdl_xs112_PreInitAdapter+0x274) [0x7fbbdbc3ec34]
10: /usr/lib/xorg/modules/drivers/fglrx_drv.so
(xdl_xs112_atiddxPreInit+0x12ad) [0x7fbbdbc4031d]
11: /usr/bin/X (InitOutput+0x83c) [0x7fbbe043287c]
12: /usr/bin/X (0x7fbbe03a0000+0x4181d) [0x7fbbe03e181d]
13: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xfd)
[0x7fbbde3f1ead]
14: /usr/bin/X (0x7fbbe03a0000+0x41cdd) [0x7fbbe03e1cdd]
Segmentation fault at address 0xffffffffe18c7332

... simply not working at all (on amd64 only maybe?)..............

[RussianNeuroMancer]

Anyone knows how to remove the watermark in the bottom right corner?

Replace control file: http://wiki.cchtml.com/index.php/Ubu...ware_Watermark
And put valid signature to signature file: http://wiki.cchtml.com/index.php/Amd_watermark

[mikejuni]

Removing watermark

Anyone knows how to remove the watermark in the bottom right corner?

[xgt001]

Some things, i have found out during normal xbmc usage:

ASIC hang has been fixed within a GL context.
HDMI Audio continues to work in the following scenario: PC -> receiver -> TV: You can switch off TV now and sound continues. You can also switch off receiver and sound continues through TV speakers.
In the scenario: PC -> TV. Sound comes back when you turn on TV after you turned it off.
VC-1 withing xvba was fixed. Artefacts were visible with 12.4
UWD_H264Level51 is still not possible.

Perhaps they change their way of publishing Changelogs now, too.

XBMC seems to work a lot better now.. as I am a laptop user, with 6370M, i barely game on either my Arch or Precise installs, but video acceleration is something which I was never satisfied with AMD.. now things definitely look better...
on other news, AMD released a Catalyst Driver specifically for Windows 8 RP on the release day with full implementation of WDDM 1.2... beat that

[d2kx]

Includes new 2D acceleration architecture. Enable with (maybe it is enabled by default... I haven't checked, but it works when setting this):

amdconfig --set-pcs-u32=DDX,ShadowPrimary,1

Supports KWin direct rendering. Force with:

KWIN_DIRECT_GL=1

Thank you RussianNeuroMancer

[fritsch]

Some things, i have found out during normal xbmc usage:

ASIC hang has been fixed within a GL context.
HDMI Audio continues to work in the following scenario: PC -> receiver -> TV: You can switch off TV now and sound continues. You can also switch off receiver and sound continues through TV speakers.
In the scenario: PC -> TV. Sound comes back when you turn on TV after you turned it off.
VC-1 withing xvba was fixed. Artefacts were visible with 12.4
UWD_H264Level51 is still not possible.

Perhaps they change their way of publishing Changelogs now, too.

[Kano]

Just began testing that driver, some facts

kernel support up to 3.3 (32 bit tested), 3.4 fails

134 legacy pci ids dropped compared to 12-4
20 new pci ids compared to 12-4
186 identical pci ids compared to 12-4