Gentoo Archives: gentoo-user

From: Andrey Moshbear <andrey.vul@×××××.com>
To: gentoo-user@l.g.o
Subject: Re: [gentoo-user] Issues with >=x11-drivers/xf86-video-ati-6.14.4: driver issue or hardware issue?
Date: Wed, 29 Aug 2012 17:04:42
Message-Id: CAFnZeoQtmxSE+DTGx03XVkcjyGJt0sn_wa9z1Fpe9S7doDbS_Q@mail.gmail.com
In Reply to: Re: [gentoo-user] Issues with >=x11-drivers/xf86-video-ati-6.14.4: driver issue or hardware issue? by Michael Scherer
1 On Tue, Jun 12, 2012 at 10:33 AM, Michael Scherer
2 <a6702894@××××××××××××××.at> wrote:
3 > On Thu, 7 Jun 2012 16:18:02 +0200
4 > Michael Scherer <a6702894@××××××××××××××.at> wrote:
5 >
6 >> On Thu, 7 Jun 2012 04:52:35 -0400
7 >> Andrey Moshbear <andrey.vul@×××××.com> wrote:
8 >>
9 >> > On Tue, Jun 5, 2012 at 11:41 AM, Michael Scherer
10 >> > <a6702894@××××××××××××××.at> wrote:
11 >> > > On Fri, 1 Jun 2012 04:11:35 -0400
12 >> > > Andrey Moshbear <andrey.vul@×××××.com> wrote:
13 >> > >
14 >> > >> On Wed, May 23, 2012 at 11:12 AM, Michael Scherer
15 >> > >> <a6702894@××××××××××××××.at> wrote:
16 >> > >> > ----- Original Message ----- From: "Andrey Moshbear"
17 >> > >> > <andrey.vul@×××××.com> To: "gentoo-user"
18 >> > >> > <gentoo-user@l.g.o> Sent: Tuesday, 22 May, 2012
19 >> > >> > 08:02 Subject: [gentoo-user] Issues with
20 >> > >> > >=x11-drivers/xf86-video-ati-6.14.4: driver issue or hardware
21 >> > >> > >issue?
22 >> > >> >
23 >> > >> >
24 >> > >> >
25 >> > >> >> Lately, I've been having some issues with segfaults when
26 >> > >> >> running startx and it's been pretty persistent.
27 >> > >> >>
28 >> > >> >> Xorg.0.log and emerge --info are available at
29 >> > >> >> https://gist.github.com/2766926 .
30 >> > >> >> Kernel config is available at https://gist.github.com/276943 .
31 >> > >> >>
32 >> > >> >> I've tried downgrading, but
33 >> > >> >> <=x11-drivers/xf86-video-ati-6.14.2 fails to compile due to
34 >> > >> >> incomplete structs.
35 >> > >> >>
36 >> > >> >> Is this more a driver or a hardware issue?
37 >> > >> >>
38 >> > >> >
39 >> > >> > first thing thing is your usage of ACCEPT_KEYWORDS="amd64
40 >> > >> > ~amd64". with ~amd64 you globally
41 >> > >> > allow all packages masked for amd64. unless you are a
42 >> > >> > developer/tester for gentoo you should remove
43 >> > >> > this keyword, because gentoo usually has good reasons to mask
44 >> > >> > some packages. if for some reason you
45 >> > >> > really need a masked package, you can do this easily only for
46 >> > >> > that package. global unmasking alone might be the reason for
47 >> > >> > half of your troubles.
48 >> > >> >
49 >> > >> > second, it is advisable to use kernel modesetting, which is
50 >> > >> > obviously not enabled. gentoo has a detailed
51 >> > >> > howto for this under
52 >> > >> > http://www.gentoo.org/doc/en/xorg-config.xml. this gives you
53 >> > >> > all necessary details.
54 >> > >> >
55 >> > >> > just a quick shot for the moment. your kernel config doesn't
56 >> > >> > under the link you give, I'd like to see that
57 >> > >> > too, and maybe /etc/X11/xorg.conf or the contents
58 >> > >> > of /etc/X11/xorg.conf.d, if you have any of these,
59 >> > >> >
60 >> > >> > regards, nichael
61 >> > >> >
62 >> > >> >
63 >> > >>
64 >> > >> Well, I added radeon drm & modesetting and the kernel is crashing
65 >> > >> right after the microcode is loaded, but without any signs of
66 >> > >> panic, be it flashing keyboard lights or kernel backtrace with
67 >> > >> register values.
68 >> > >>
69 >> > >> Also, the segfault was caused by the DRI code interpreting a DRI
70 >> > >> opcode as a pointer, hence 0xa4 or similar in the back trace.
71 >> > >>
72 >> > >> --
73 >> > >> m0shbear
74 >> > >>
75 >> > >
76 >> > > in any case, I'd need some more information:
77 >> > > the link to your .config is broken, I only get an empty page.
78 >> > > if the kernel is involved, output of dmesg and rc.log would
79 >> > > be needed to.
80 >> > >
81 >> > > by the way, did you anything with your ACCEPT_KEYWORDS?
82 >> > > if not, the best thing would be set it simply ="" (and
83 >> > > afterwards --update world, --depclean and revdep-rebuild),
84 >> > > but that's up to you.
85 >> > >
86 >> > > please send me the requested information, otherwise it's
87 >> > > difficult to find out what's wrong.
88 >> > >
89 >> >
90 >> > Also, I've noticed that that radeon.ko autodetection configures it
91 >> > as a RS780, whereas, according to motherboard documentation, it
92 >> > should be RS880 (780G vs 785G).
93 >> > This may be a reason for crashing, since the video card dies due to
94 >> > being supplied bad microcode.
95 >> >
96 >>
97 >> according to the gentoo manual, under "device drivers-graphics-
98 >> support for framebufferdevices" you should uncheck
99 >> -everything- except "enable video mode handling helpers".
100 >> instead of uvesafb in the kernel parameters you may use
101 >> "radeondrmfb", if anything at all.
102 >>
103 >> from the Xorg radeon page:
104 >> "First of all check that you don't load radeonfb, uvesafb or vesafb
105 >> module. This includes no vga parameters for kernel when using KMS.
106 >> Console is provided by fbcon and radeondrmfb frame buffer console. So
107 >> it is best to make sure that fbcon module is loaded"
108 >>
109 >> disabling vesa also should rid you of uvesa messages in Xorg.0.log.
110 >>
111 >> the rs880 and configuring as rs780 is indeed probably at the root
112 >> of your troubles. seems that you got the wrong firmware. I have
113 >> found other postings with the exact same problem, but as yet
114 >> no clear solution. I'll see what I can find and let you know.
115 >>
116 >>
117 >
118 > if you haven't solved your problem already, there are two
119 > possibilities you coud try:
120 > one is using the Radeon R600-family RLC microcode: radeon/R600_rlc.bin
121 > (without something else). this should include the rs880.
122 > another option might be to download firmware-linux-nonfree from
123 > http://packages.debian.org/de/squeeze/firmware-linux-nonfree
124 > some people seem to have used it successfully, but I don't
125 > know if it works on gentoo.
126 >
127 > sorry that I can only guess on this, not having an rs880 myself.
128 > isn't someone on gentoo users with a similar configuration who
129 > could give you better advice?
130 >
131
132 Interesting updates:
133 1) RS780 is the proper firmwareset, as evidenced by knoppix.
134 2) In 3.5.2, while it appears to hang after loading firmware,
135 <Alt><SysRq>R keeps the process flowing along. I will say this -
136 switching between vt and X in the kms radeon driver is a *** of a lot
137 faster than in the nvidia blob.