Date: Tue, 19 May 2009 14:18:17 -0500 From: Robert Noland <rnoland@FreeBSD.org> To: Tijl Coosemans <tijl@ulyssis.org> Cc: freebsd-emulation@freebsd.org Subject: Re: [new port] graphics/linux-dri74 Message-ID: <1242760698.1752.21.camel@balrog.2hip.net> In-Reply-To: <200905192012.06401.tijl@ulyssis.org> References: <92596693@bb.ipt.ru> <20090515145252.GA69488@yamagi.org> <20090519165039.GA12505@yamagi.org> <200905192012.06401.tijl@ulyssis.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--=-1+LxFuJAo2WPGi6ytpd6 Content-Type: text/plain Content-Transfer-Encoding: quoted-printable On Tue, 2009-05-19 at 20:12 +0200, Tijl Coosemans wrote: > On Tuesday 19 May 2009 18:50:39 Yamagi Burmeister wrote: > > Okay, I investigated further. First I tried a FreeBSD/i386 version of > > ioQuake3 on a FreeBSD/amd64 host. After some minor and unrelated > > problems it worked. > >=20 > > After that I tried to debug the linux-dri74. Reading the source and > > doing a lot test runs I realized, that there are in fact two > > problems. Every ten starts or so the userland side of drm crashes: > >=20 > > yamagi@saya:ttyp3 ~: /usr/compat/linux/usr/bin/glxinfo > > name of display: :0.0 > > libGL: XF86DRIGetClientDriverName: 5.3.0 r300 (screen 0) > > libGL: OpenDriver: trying /usr/lib/dri/tls/r300_dri.so > > libGL: OpenDriver: trying /usr/lib/dri/r300_dri.so > > drmOpenDevice: node name is /dev/dri/card0 > > drmOpenDevice: open result is 4, (OK) > > DRM_IOCTL_VERSION: Bad address > > Segmentation fault (core dumped) > >=20 > > The other times the userland side of drm is able to open the > > drm-device bit is unable to initalize it. The error message is > > somewhat missleading: > >=20 > > yamagi@saya:ttyp3 ~: /usr/compat/linux/usr/bin/glxinfo > > name of display: :0.0 > > libGL: XF86DRIGetClientDriverName: 5.3.0 r300 (screen 0) > > libGL: OpenDriver: trying /usr/lib/dri/tls/r300_dri.so > > libGL: OpenDriver: trying /usr/lib/dri/r300_dri.so > > drmOpenDevice: node name is /dev/dri/card0 > > drmOpenDevice: open result is 4, (OK) > > drmOpenByBusid: Searching for BusID pci:0000:01:00.0 > > drmOpenDevice: node name is /dev/dri/card0 > > drmOpenDevice: open result is 4, (OK) > > drmOpenByBusid: drmOpenMinor returns 4 > > drmOpenByBusid: drmGetBusid reports (null) > > drmOpenDevice: node name is /dev/dri/card1 > > drmOpenByBusid: drmOpenMinor returns -1 > > drmOpenDevice: node name is /dev/dri/card2 > > drmOpenByBusid: drmOpenMinor returns -1 > > drmOpenDevice: node name is /dev/dri/card3 > > drmOpenByBusid: drmOpenMinor returns -1 > > drmOpenDevice: node name is /dev/dri/card4 > > drmOpenByBusid: drmOpenMinor returns -1 > > [..] > > drmOpenDevice: node name is /dev/dri/card14 > > drmOpenByBusid: drmOpenMinor returns -1 > > libGL error: drmOpenOnce failed (Operation not permitted) > > libGL error: reverting to software direct rendering > > libGL: OpenDriver: trying /usr/lib/dri/tls/swrast_dri.so > > libGL: OpenDriver: trying /usr/lib/dri/swrast_dri.so > > display: :0 screen: 0 > > direct rendering: Yes > >=20 > > So I added debug printf() into the code. I stepped through it. After > > some hours of intesive debugging im 99% sure, that both of the above > > failures of linux-drm74 are originating at the kernel side. It's most > > likely NOT a problem of the userland side and therefor not a problem > > of the port. > >=20 > > I'll investigate further at the kernel side, but that'll need some > > time. So, while it's the fault linux-drm74, what about a little > > warning message in the install message of the port? Just to be sure, > > that other users experiencing the same problem know, that it is not a > > bug of linux-dr74. >=20 > I noticed you use the r300 driver so maybe this patch helps: > http://people.freebsd.org/~rnoland/drm_radeon-copyin-fix-try2.patch This patch, while it may be useful isn't related to the issue. > Robert Noland (CCed) probably knows more about this though. > http://lists.freebsd.org/pipermail/freebsd-emulation/2009-May/005995.html It looks like the linux version of drmOpen is behaving badly. Possibly, it needs linproc mounted. drmOpen appears to be returning file descriptor 4, which may be being mis-interpreted, but in any case it really shouldn't need keep looking for more cards. robert. --=20 Robert Noland <rnoland@FreeBSD.org> FreeBSD --=-1+LxFuJAo2WPGi6ytpd6 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.11 (FreeBSD) iEYEABECAAYFAkoTBfkACgkQM4TrQ4qfROPnfACfUvIoP/Ds4KVVVSv0lNWoh39B bOMAn3d6qDu8VEB1+okhFEMS0+W6M8OM =VxOQ -----END PGP SIGNATURE----- --=-1+LxFuJAo2WPGi6ytpd6--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1242760698.1752.21.camel>