From mboxrd@z Thu Jan 1 00:00:00 1970 From: younes.m-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org Subject: Re: nouveaufb problems with NV18 Date: Fri, 01 Jul 2011 17:13:49 +0000 Message-ID: <90e6ba6135a819446804a705242e@google.com> References: <20110701163521.GA3162@joi.lan> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============1110918829==" Return-path: In-Reply-To: <20110701163521.GA3162-OI9uyE9O0yo@public.gmane.org> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: nouveau-bounces+gcfxn-nouveau=m.gmane.org-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org Errors-To: nouveau-bounces+gcfxn-nouveau=m.gmane.org-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org To: Marcin Slusarz Cc: nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org, George Spelvin List-Id: nouveau.vger.kernel.org --===============1110918829== Content-Type: multipart/alternative; boundary=90e6ba6135a819445204a705242b --90e6ba6135a819445204a705242b Content-Type: text/plain; charset=ISO-8859-1; format=flowed; delsp=yes On , Marcin Slusarz wrote: > On Fri, Jul 01, 2011 at 01:16:49AM -0400, George Spelvin wrote: > > I mentioned these problems back in March, but went back to the nv driver > > and ignored it. Since I need to reboot for another reason, I gave > nouveau a try again, > > > > This time, I'm using the kernel git tip, off of 3.0-rc5: > > 181888bfeea73b2ec756b83fd590fd3a62d4171d: drm/nv50/dp: fix hack to work > for macbooks booted via EFI > > > > The X server is Debian's xserver-xorg-video-nouveau, version > > 1:0.0.16+git20110411+8378443-1+b1 > > > > The good news is that my screen "tearing" problem during system activity > > appears to be fixed! > > Hooray. But I still have two other problems I mentioned at the time: > > > > 1) The text console, although connected via DVI, appears to be about > > half a line of text low. Tall characters (^ and $) on the top line > > have their top row of pixels "stretched" upward, apparently replicated > > quite a few times. > > The bottom row of characters is only partially visible. > Screenshot might be useful. > > > > 2) I can't start X. The X server fails to initialize with: > > [ 6151.976] (II) NOUVEAU(0): Opened GPU channel 1 > > [ 6151.976] (II) NOUVEAU(0): [DRI2] Setup complete > > [ 6151.976] (II) NOUVEAU(0): [DRI2] DRI driver: nouveau_vieux > Can you uninstall 3d driver and check how will it behave? > > [ 6151.977] (EE) NOUVEAU(0): Error allocating scanout buffer: 0 > > [ 6151.977] > > Fatal server error: > > [ 6151.977] AddScreen/ScreenInit failed for driver 0 > > [ 6151.977] > > > > If I can get problem 2 fixed, problem 1 will become less important. > > > > As mentioned in my old email, it's a uniprocessor Athlon XP system, > with a > > Geforce 4 MX video adapter: > > 04:00.0 VGA compatible controller [0300]: nVidia Corporation NV18 > [GeForce4 MX - nForce GPU] [10de:01f0] (rev a3) > FYI, nouveau finds it as NV1F. > > The display is a 1920x1080 LCD, connected via DVI. No other outputs are > connected. > 1920x1080x4 is almost 8MB and you have only 32MB of VRAM. It might be a > problem with memory allocation. > You could verify it by applying this patch to xf86-video-nouveau and > checking xorg.log: > diff --git a/src/nv_accel_common.cb/src/nv_accel_common.c > index 735f47f..1eac357 100644 > --- a/src/nv_accel_common.c > +++ b/src/nv_accel_common.c > @@ -116,8 +116,10 @@ nouveau_allocate_surface(ScrnInfoPtr scrn, int > width, int height, int bpp, > ret = nouveau_bo_new_tile(pNv->dev, flags, 0, *pitch * height, > tile_mode, tile_flags, bo); > - if (ret) > + if (ret) { > + xf86DrvMsg(scrn->scrnIndex, X_ERROR, "nouveau_bo_new_tile > returned: %d, %s\n", ret, strerror(ret)); > return FALSE; > + } > return TRUE; > } > - > Lowering bpp/resolution for framebuffer and/or X might help. At least > temporarily. Likely low on VRAM. I don't see why it would be able to create the 1st frame buffer for fbcon but not a 2nd for X unless it was tight on memory. --90e6ba6135a819445204a705242b Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable On , Marcin Slusarz <marcin.slusarz-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> On Fr= i, Jul 01, 2011 at 01:16:49AM -0400, George Spelvin wrote:
>
= > > I mentioned these problems back in March, but went back to the nv= driver
>
> > and ignored it. =A0Since I need to reboot= for another reason, I gave nouveau a try again,
>
> ><= br />>
> > This time, I'm using the kernel git tip, off = of 3.0-rc5:
>
> > 181888bfeea73b2ec756b83fd590fd3a62d41= 71d: drm/nv50/dp: fix hack to work for macbooks booted via EFI
> > >
>
> > The X server is Debian's xserve= r-xorg-video-nouveau, version
>
> > 1:0.0.16+git2011041= 1+8378443-1+b1
>
> >
>
> > The good= news is that my screen "tearing" problem during system activity<= br />>
> > appears to be fixed!
>
> > Ho= oray. =A0But I still have two other problems I mentioned at the time:
= >
> >
>
> > 1) The text console, althoug= h connected via DVI, appears to be about
>
> > =A0 =A0h= alf a line of text low. =A0Tall characters (^ and $) on the top line
&= gt;
> > =A0 =A0have their top row of pixels "stretched&quo= t; upward, apparently replicated
>
> > =A0 =A0quite a f= ew times.
>
> > =A0 =A0The bottom row of characters is = only partially visible.
>
>
>
> Screensh= ot might be useful.
>
>
>
> >
&g= t;
> > 2) I can't start X. =A0The X server fails to initial= ize with:
>
> > [ =A06151.976] (II) NOUVEAU(0): Opened = GPU channel 1
>
> > [ =A06151.976] (II) NOUVEAU(0): [DR= I2] Setup complete
>
> > [ =A06151.976] (II) NOUVEAU(0)= : [DRI2] =A0 DRI driver: nouveau_vieux
>
>
>
> Can you uninstall 3d driver and check how will it behave?
> =
>
>
> > [ =A06151.977] (EE) NOUVEAU(0): Error= allocating scanout buffer: 0
>
> > [ =A06151.977]
>
> > Fatal server error:
>
> > [ =A061= 51.977] AddScreen/ScreenInit failed for driver 0
>
> > = [ =A06151.977]
>
> >
>
> > If I can= get problem 2 fixed, problem 1 will become less important.
>
> >
>
> > As mentioned in my old email, it's= a uniprocessor Athlon XP system, with a
>
> > Geforce = 4 MX video adapter:
>
> > 04:00.0 VGA compatible contro= ller [0300]: nVidia Corporation NV18 [GeForce4 MX - nForce GPU] [10de:01f0]= (rev a3)
>
>
>
> FYI, nouveau finds it = as NV1F.
>
>
>
> > The display is a 1= 920x1080 LCD, connected via DVI. =A0No other outputs are connected.
&g= t;
>
>
> 1920x1080x4 is almost 8MB and you have = only 32MB of VRAM. It might be a problem with memory allocation.
> =
> You could verify it by applying this patch to xf86-video-nouveau= and checking xorg.log:
>
>
>
> diff --g= it a/src/nv_accel_common.c b/src/nv_accel_common.c
>
> ind= ex 735f47f..1eac357 100644
>
> --- a/src/nv_accel_common.c=
>
> +++ b/src/nv_accel_common.c
>
> @@ -= 116,8 +116,10 @@ nouveau_allocate_surface(ScrnInfoPtr scrn, int width, int = height, int bpp,
>
>
>
> =A0 =A0 =A0 = =A0ret =3D nouveau_bo_new_tile(pNv->dev, flags, 0, *pitch * height,
>
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = =A0 =A0 =A0tile_mode, tile_flags, bo);
>
> - =A0 =A0 =A0 i= f (ret)
>
> + =A0 =A0 =A0 if (ret) {
>
> = + =A0 =A0 =A0 =A0 =A0 =A0 =A0 xf86DrvMsg(scrn->scrnIndex, X_ERROR, "= ;nouveau_bo_new_tile returned: %d, %s\n", ret, strerror(ret));
&g= t;
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0return FALSE;
>
> + =A0 =A0 =A0 }
>
>
>
> =A0 =A0 = =A0 =A0return TRUE;
>
> =A0}
>
>
&g= t;
> -
>
>
>
> Lowering bpp/res= olution for framebuffer and/or X might help. At least temporarily.
Likely low on VRAM. I don't see why it would be able to create the 1= st frame buffer for fbcon but not a 2nd for X unless it was tight on memory= . --90e6ba6135a819445204a705242b-- --===============1110918829== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Nouveau mailing list Nouveau-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org http://lists.freedesktop.org/mailman/listinfo/nouveau --===============1110918829==--