10.4 Released

[etreus]

@etreus

Use my script or update the packages dir after --extract with

http://phorogit.com/index.php?p=fglr...ing.git&dl=zip

if you prefer manual way. My script adds kernel .33/.34 support on the fly too. Did not test on karmic however - you can do

I fix with

binary-install/$(PKG_control)::
#execstack -c debian/$(PKG_control)/$(DRIDIR)/bin/amdcccle

on packages/Ubuntu/dists/karmic/rules

before I extract it with

sh ati-driver-installer-10-4-x86.x86_64.run --extract

[etreus]

@etreus

Use my script or update the packages dir after --extract with

http://phorogit.com/index.php?p=fglr...ing.git&dl=zip

if you prefer manual way. My script adds kernel .33/.34 support on the fly too. Did not test on karmic however - you can do

1) extract the data:
sh ati-driver-installer-10-4-x86.x86_64.run --extract ati_10.04

2) edit the file:
vi ati_10.04/packages/Ubuntu/dists/karmic/rules

edit the last line

binary-install/$(PKG_control)::
        #execstack -c debian/$(PKG_control)/$(DRIDIR)/bin/amdcccle

3) from dir ati_10.04:
./ati-installer.sh 10.04 --buildpkg Ubuntu/karmic

[etreus]

1) extract the data:
sh ati-driver-installer-10-4-x86.x86_64.run --extract ati_10.04

2) edit the file:
vi ati_10.04/packages/Ubuntu/dists/karmic/rules

edit the last line

binary-install/$(PKG_control)::
        #execstack -c debian/$(PKG_control)/$(DRIDIR)/bin/amdcccle

3) from dir ati_10.04:
./ati-installer.sh 10.04 --buildpkg Ubuntu/karmic

[cutterjohn]

1) extract the data:
sh ati-driver-installer-10-4-x86.x86_64.run --extract ati_10.04

2) edit the file:
vi ati_10.04/packages/Ubuntu/dists/karmic/rules

edit the last line

binary-install/$(PKG_control)::
        #execstack -c debian/$(PKG_control)/$(DRIDIR)/bin/amdcccle

3) from dir ati_10.04:
./ati-installer.sh 10.04 --buildpkg Ubuntu/karmic

Nope. Doesn't work.

Forgot to check the directory created in /tmp first though, so maybe I'll invest a few minutes at some time to see what exactly it's trying to do...

[cutterjohn]

Nope. Doesn't work.

Forgot to check the directory created in /tmp first though, so maybe I'll invest a few minutes at some time to see what exactly it's trying to do...

OK. I gave it one last shot, and noticed this time that it was apparently trying to create a file in tmp, pkg_build.out but there was already one left over from when I tried an earlier build with the owner of root.

Removing that pkg_build.out allowed it to build ok this time.

[cutterjohn]

2 more comments,
Ubuntu 9.10 x86-64
mobility 4850
catalyst 10.4 as built via above rules script edit.

My LCD brightness control has gone completely wonky from just semi-f'ed up. (Before 10.4 the 2nd setting from none in the brightness popup actually seemed to be the lowest brightness setting while the lower 2 became progressively brighter, i.e. 1 up from none was brighter than 2 up, and brightness all the way "down"(no bar) was brighter than 1 "up". Now my brightness control just flops all over the place, from nothing to what looks like about 5 then to 2, then to 4, etc... best way to describe I guess would be highlighted segments of the brightness popup feedback indicator...)

2. "Failed to execute child process "amdxdg-su" (No such file or directory)"
CCC is back to shipping w/o AMD's little superuser script or whatever it was again. It's been like this for several releases now, and I just remember to try it again since I kept my own gksudo startup workaround from the last time...

[Kano]

Well with the hotfix mentioned by etreus you will only have got a build fix, that's not that same what you get when you update the complete packageing which i do in my script. For testing my script has a -u switch that will always fetch the packaging from the phorogit zip file. Usually i embedd the needed patches inside the script so that download is not needed for tested driver releases. Its needed to do that when the packages change - which definitely happens from time to time.

[mudd1]

I just love the "Known Issues" section in their release notes. If this heading doesn't lie, I think it's safe to say that AMD just doesn't know nearly as much as they easily could know: http://ati.cchtml.com/buglist.cgi?bu...severity=major

[haplo602]

so it's X 1.7 for me again ... I could have waited the month instead of downgrading to 1.6 for 10.3 to work :-(((

ok happy upgrading again today ...

[wall]

I'm on gentoo, 10.4 works great here, although I had to revert back to xorg-server 1.6.5 because of some "resVgaShared" issue.

If there's a patch to get it working on 1.7, I'm interested. Mind you, Apart from 1.7 going stable, I'm not even sure what the benefits with 1.7 are.

Re, the minimization/resize delay. The lag seems shorter with every release.