Date: Thu, 09 Sep 2004 10:37:44 +0400 From: Vladimir Grebenschikov <vova@fbsd.ru> To: Eric Anholt <eta@lclark.edu> Cc: "current@freebsd.org" <current@freebsd.org> Subject: Re: ATI Radeon LY Mobility M6: DRM does not work - locking issue ? Message-ID: <1094711864.1043.9.camel@localhost> In-Reply-To: <1094686987.871.10.camel@leguin> References: <1094502674.2668.4.camel@localhost> <1094631615.860.7.camel@leguin> <1094634253.2172.13.camel@localhost> <1094686987.871.10.camel@leguin>
next in thread | previous in thread | raw e-mail | index | archive | help
=F7 =D3=D2, 08/09/2004 =D7 16:43 -0700, Eric Anholt =D0=C9=DB=C5=D4: > On Wed, 2004-09-08 at 05:23, Jose M Rodriguez wrote: > > On Wed, 08 Sep 2004 13:04:13 +0400, Vladimir Grebenschikov <vova@fbsd.r= u> =20 > > wrote: > >=20 > > > =F7 =D3=D2, 08/09/2004 =D7 01:20 -0700, Eric Anholt =D0=C9=DB=C5=D4: > > >> On Mon, 2004-09-06 at 13:31, Vladimir Grebenschikov wrote: > > >> > Hi > > >> > > > >> > 6-CURRENT > > >> > > > >> > from dmesg: > > >> > > > >> > drm0: <ATI Radeon LY Mobility M6> port 0x3000-0x30ff mem > > >> > 0xd0100000-0xd010ffff,0xd8000000-0xdfffffff irq 9 at device 0.0 on= =20 > > >> pci1 > > >> > info: [drm] Initialized radeon 1.11.0 20020828 on minor 0 > > >> > error: [drm:pid2477:radeon_cp_init] *ERROR* radeon_cp_init called > > >> > without lock held > > >> > error: [drm:pid2477:radeon_unlock] *ERROR* Process 2477 using kern= el > > >> > context 0 > > >> > > >> If you look in your dmesg, does agp initialization come before or af= ter > > >> the drm? This sure looks like the symptoms of agp initialization co= ming > > >> after (or not at all), but if you say your agp is loaded and attache= d, > > >> I'm not sure how that would happen. > > > > > > dmesg: > > > > > > drm0: <ATI Radeon LY Mobility M6> port 0x3000-0x30ff mem > > > 0xd0100000-0xd010ffff,0xd8000000-0xdfffffff irq 9 at device 0.0 on pc= i1 > > > info: [drm] AGP at 0xe0000000 256MB > > > info: [drm] Initialized radeon 1.11.0 20020828 on minor 0 > > > > > > I have solved problem, it need to make depth 16 or there is no enough > > > memory to serve direct rendering on 1400x1050 screen with 24 bpp. > > > > >=20 > > Get an X -configure and check driver options. > > You may need: > > Options "AGPSize" "256" # AGP Aperture size. MUST fit in bios setting= s >=20 > This won't affect his problem, since only command buffers (and maybe > textures -- I don't remember the status of our R100 driver) get > allocated from AGP, and not the front/back/depth. It will also wire > 256MB of memory for essentially no purpose at all. Yes, it does not helps, with 256 it makes mess on X start, and freezes console, with 64 it allocates 64 Mb, mostly for textures: (II) RADEON(0): [agp] Mode 0x1f000217 [AGP 0x0000/0x0000; Card 0x1002/0x4c59] (II) RADEON(0): [agp] 65536 kB allocated with handle 0xc1e9c740 (II) RADEON(0): [agp] ring handle =3D 0xe0000000 (II) RADEON(0): [agp] Ring mapped at 0x29422000 (II) RADEON(0): [agp] ring read ptr handle =3D 0xe0101000 (II) RADEON(0): [agp] Ring read ptr mapped at 0x282cf000 (II) RADEON(0): [agp] vertex/indirect buffers handle =3D 0xe0102000 (II) RADEON(0): [agp] Vertex/indirect buffers mapped at 0x29523000 (II) RADEON(0): [agp] GART texture map handle =3D 0xe0302000 (II) RADEON(0): [agp] GART Texture map mapped at 0x29723000 (II) RADEON(0): [drm] register handle =3D 0xd0100000 (II) RADEON(0): [dri] Visual configs initialized (II) RADEON(0): CP in BM mode (II) RADEON(0): Using 64 MB GART aperture (II) RADEON(0): Using 1 MB for the ring buffer (II) RADEON(0): Using 2 MB for vertex/indirect buffers (II) RADEON(0): Using 61 MB for GART textures (II) RADEON(0): Memory manager initialized to (0,0) (1408,5957) (II) RADEON(0): Reserved area from (0,1050) to (1408,1052) (II) RADEON(0): Largest offscreen area available: 1408 x 4905 When with default value it is: (II) RADEON(0): [agp] Mode 0x1f000217 [AGP 0x0000/0x0000; Card 0x1002/0x4c59] (II) RADEON(0): [agp] 8192 kB allocated with handle 0xc1bba000 (II) RADEON(0): [agp] ring handle =3D 0xe0000000 (II) RADEON(0): [agp] Ring mapped at 0x29422000 (II) RADEON(0): [agp] ring read ptr handle =3D 0xe0101000 (II) RADEON(0): [agp] Ring read ptr mapped at 0x282cf000 (II) RADEON(0): [agp] vertex/indirect buffers handle =3D 0xe0102000 (II) RADEON(0): [agp] Vertex/indirect buffers mapped at 0x29523000 (II) RADEON(0): [agp] GART texture map handle =3D 0xe0302000 (II) RADEON(0): [agp] GART Texture map mapped at 0x29723000 (II) RADEON(0): [drm] register handle =3D 0xd0100000 (II) RADEON(0): [drm] register handle =3D 0xd0100000 (II) RADEON(0): [dri] Visual configs initialized (II) RADEON(0): CP in BM mode (II) RADEON(0): Using 8 MB GART aperture (II) RADEON(0): Using 1 MB for the ring buffer (II) RADEON(0): Using 2 MB for vertex/indirect buffers (II) RADEON(0): Using 5 MB for GART textures (II) RADEON(0): Memory manager initialized to (0,0) (1408,5957) (II) RADEON(0): Reserved area from (0,1050) to (1408,1052) (II) RADEON(0): Largest offscreen area available: 1408 x 4905 > > Options "AGPMode" "4" # Set AGP Speed DANGEROUS but very effective >=20 > I would be interested in reports of "very effective." The effectiveness > that I've heard from anyone so far is "not measurable" to 1-2%. Seems to be enabled successful (both mode 4 and fast-write): (**) RADEON(0): Using AGP 4x mode (**) RADEON(0): Enabling AGP Fast Write (II) RADEON(0): Depth moves disabled by default But can't see any difference with glxgears as fps metter. it is gives about 470 fps with default window size on=20 ATI M6 16Mb + PentiumM 1700MHz --=20 Vladimir B. Grebenchikov vova@fbsd.ru
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1094711864.1043.9.camel>