Most recent kernel where this bug did not occur:2.6.21 Distribution:gentoo Hardware Environment:HP nx6125 BIOS version F.11 Software Environment:Gnome 2.18, ati-driver 8.40.4 Problem Description:If you enable "support for frame buffer devices", "Vesa vga graphics support" and "ati radeon display support" under "graphics support" so you lost fan cantrol on your hp nx6125, that means the fan doens't stop working. Steps to reproduce:Enable "support for frame buffer devices", "Vesa vga graphics support" and "ati radeon display support" under "graphics support" on your hp nx6125 and after the next boot the fan will not switched off.
Reply-To: akpm@linux-foundation.org On Sat, 13 Oct 2007 02:37:35 -0700 (PDT) bugme-daemon@bugzilla.kernel.org wrote: > http://bugzilla.kernel.org/show_bug.cgi?id=9155 > > Summary: No fan control with ATI Radeon display support on HP > nx6125. > Product: Drivers > Version: 2.5 > KernelVersion: 2.6.22, 2.6.23 > Platform: All > OS/Version: Linux > Tree: Mainline > Status: NEW > Severity: low > Priority: P1 > Component: Video(Other) > AssignedTo: drivers_video-other@kernel-bugs.osdl.org > ReportedBy: daniel.wuerfel@gmx.net > > > Most recent kernel where this bug did not occur:2.6.21 > Distribution:gentoo > Hardware Environment:HP nx6125 BIOS version F.11 > Software Environment:Gnome 2.18, ati-driver 8.40.4 > Problem Description:If you enable "support for frame buffer devices", "Vesa > vga > graphics support" and "ati radeon display support" under "graphics support" > so > you lost fan cantrol on your hp nx6125, that means the fan doens't stop > working. > > Steps to reproduce:Enable "support for frame buffer devices", "Vesa vga > graphics support" and "ati radeon display support" under "graphics support" > on > your hp nx6125 and after the next boot the fan will not switched off. > Does anyone know whether this is likely to be a radeon driver bug or an acpi bug or something else??
On Saturday, 13 October 2007 19:53, Andrew Morton wrote: > On Sat, 13 Oct 2007 02:37:35 -0700 (PDT) bugme-daemon@bugzilla.kernel.org > wrote: > > > Most recent kernel where this bug did not occur:2.6.21 > > Distribution:gentoo > > Hardware Environment:HP nx6125 BIOS version F.11 > > Software Environment:Gnome 2.18, ati-driver 8.40.4 Looks like the binary ATI driver is used. > > Problem Description:If you enable "support for frame buffer devices", "Vesa > vga > > graphics support" and "ati radeon display support" under "graphics support" > so > > you lost fan cantrol on your hp nx6125, that means the fan doens't stop > > working. > > > > Steps to reproduce:Enable "support for frame buffer devices", "Vesa vga > > graphics support" and "ati radeon display support" under "graphics support" > on > > your hp nx6125 and after the next boot the fan will not switched off. > > > > Does anyone know whether this is likely to be a radeon driver bug or an > acpi bug or something else?? It looks like a thermal issue to me, so I'd say ACPI or the ATI driver (or the combination of the two).
Hi there, at first great respect for your fast answers. I don't think, that the binary ati driver is the problem, because: 1. with switched of "ati radeon display support" in the kernel, the fan control works correctly (and i still using the ati-driver) 2. the fan starts spinning at the boot screen even if the radeon display driver was loaded (and the ati driver is not just loaded yet). Here you see the difference between 2.6.21, when it all works correctly, and 2.6.22 and 2.6.23: the screen resolution of the boot screen changed to the native resolution 1024x768 (in 2.6.21 it still the default terminal resolution) Am Samstag, den 13.10.2007, 20:30 +0200 schrieb Rafael J. Wysocki: > On Saturday, 13 October 2007 19:53, Andrew Morton wrote: > > On Sat, 13 Oct 2007 02:37:35 -0700 (PDT) bugme-daemon@bugzilla.kernel.org > wrote: > > > > > Most recent kernel where this bug did not occur:2.6.21 > > > Distribution:gentoo > > > Hardware Environment:HP nx6125 BIOS version F.11 > > > Software Environment:Gnome 2.18, ati-driver 8.40.4 > > Looks like the binary ATI driver is used. > > > > Problem Description:If you enable "support for frame buffer devices", > "Vesa vga > > > graphics support" and "ati radeon display support" under "graphics > support" so > > > you lost fan cantrol on your hp nx6125, that means the fan doens't stop > > > working. > > > > > > Steps to reproduce:Enable "support for frame buffer devices", "Vesa vga > > > graphics support" and "ati radeon display support" under "graphics > support" on > > > your hp nx6125 and after the next boot the fan will not switched off. > > > > > > > Does anyone know whether this is likely to be a radeon driver bug or an > > acpi bug or something else?? > > It looks like a thermal issue to me, so I'd say ACPI or the ATI driver (or > the > combination of the two).
On Sunday, 14 October 2007 09:26, Daniel W
Am Sonntag, den 14.10.2007, 16:02 +0200 schrieb Rafael J. Wysocki: > On Sunday, 14 October 2007 09:26, Daniel Würfel wrote: > > Hi there, > > > > at first great respect for your fast answers. > > I don't think, that the binary ati driver is the problem, because: > > 1. with switched of "ati radeon display support" in the kernel, the fan > > control works correctly (and i still using the ati-driver) > > How do you know, however, that the "ati radeon display support" in the kernel > does not interfere with the ATI driver in a destructive way? I think, but im not sure, that the ati driver is loaded, if X was started. But at this moment X was not started. > > > 2. the fan starts spinning at the boot screen even if the radeon display > > driver was loaded (and the ati driver is not just loaded yet). > > I'm not sure what you mean exactly. Can you clarify, please?dann Well, i see it looks a little bit confused what i wrote. What i mean is, that the fan at this time starts to rotate, when the "radeon display support" is loaded at boot time. Normally the fan should start to work if a special temperature threshold is reached. > > > Here you see the difference between 2.6.21, when it all works correctly, > and > > 2.6.22 and 2.6.23: the screen resolution of the boot screen changed to > > the native resolution 1024x768 (in 2.6.21 it still the default terminal > > resolution) > > This only means that the kernel's radeon driver has changed. > > It's difficult to say what's going wrong. > > What's the CPU load when the fan is spinning? > > Greetings, > Rafael The rest of the systems works fine. I notice no anomaly. Greetings, Daniel
On Sunday, 14 October 2007 16:18, Daniel Würfel wrote: > Am Sonntag, den 14.10.2007, 16:02 +0200 schrieb Rafael J. Wysocki: > > On Sunday, 14 October 2007 09:26, Daniel Würfel wrote: > > > Hi there, > > > > > > at first great respect for your fast answers. > > > I don't think, that the binary ati driver is the problem, because: > > > 1. with switched of "ati radeon display support" in the kernel, the fan > > > control works correctly (and i still using the ati-driver) > > > > How do you know, however, that the "ati radeon display support" in the > kernel > > does not interfere with the ATI driver in a destructive way? > > I think, but im not sure, that the ati driver is loaded, if X was > started. But at this moment X was not started. Well, the kernel module part is probably loaded earlier. > > > 2. the fan starts spinning at the boot screen even if the radeon display > > > driver was loaded (and the ati driver is not just loaded yet). > > > > I'm not sure what you mean exactly. Can you clarify, please?dann > > Well, i see it looks a little bit confused what i wrote. What i mean is, > that the fan at this time starts to rotate, when the "radeon display > support" is loaded at boot time. I see. Can you please check: - What happens if you don't use the ATI driver at all (please unload its kernel module before loading the kernel's radeon one)? - Does the loading of the radeon module trigger the fan automatically? - What happens if the fan module is not loaded at that time?
Am Sonntag, den 14.10.2007, 21:55 +0200 schrieb Rafael J. Wysocki: > On Sunday, 14 October 2007 16:18, Daniel Würfel wrote: > > Am Sonntag, den 14.10.2007, 16:02 +0200 schrieb Rafael J. Wysocki: > > > On Sunday, 14 October 2007 09:26, Daniel Würfel wrote: > > > > Hi there, > > > > > > > > at first great respect for your fast answers. > > > > I don't think, that the binary ati driver is the problem, because: > > > > 1. with switched of "ati radeon display support" in the kernel, the fan > > > > control works correctly (and i still using the ati-driver) > > > > > > How do you know, however, that the "ati radeon display support" in the > kernel > > > does not interfere with the ATI driver in a destructive way? > > > > I think, but im not sure, that the ati driver is loaded, if X was > > started. But at this moment X was not started. > > Well, the kernel module part is probably loaded earlier. > > > > > 2. the fan starts spinning at the boot screen even if the radeon > display > > > > driver was loaded (and the ati driver is not just loaded yet). > > > > > > I'm not sure what you mean exactly. Can you clarify, please?dann > > > > Well, i see it looks a little bit confused what i wrote. What i mean is, > > that the fan at this time starts to rotate, when the "radeon display > > support" is loaded at boot time. > > I see. > > Can you please check: > - What happens if you don't use the ATI driver at all (please unload its > kernel > module before loading the kernel's radeon one)? > - Does the loading of the radeon module trigger the fan automatically? > - What happens if the fan module is not loaded at that time? I think i got the problem: I made some test. The problem is the module radeonfb. If i compile radeonfb as a module there is no problem with the fan control. But if i compile it into the kernel, than fan control will be destroyed and the fan doesn't stop working from the beginning of the system start. The other modules (fglrx, agp_gart) have no influence to the fan behaviour.
On Monday, 15 October 2007 19:12, Daniel Würfel wrote: > Am Sonntag, den 14.10.2007, 21:55 +0200 schrieb Rafael J. Wysocki: > > On Sunday, 14 October 2007 16:18, Daniel Würfel wrote: > > > Am Sonntag, den 14.10.2007, 16:02 +0200 schrieb Rafael J. Wysocki: > > > > On Sunday, 14 October 2007 09:26, Daniel Würfel wrote: > > > > > Hi there, > > > > > > > > > > at first great respect for your fast answers. > > > > > I don't think, that the binary ati driver is the problem, because: > > > > > 1. with switched of "ati radeon display support" in the kernel, the > fan > > > > > control works correctly (and i still using the ati-driver) > > > > > > > > How do you know, however, that the "ati radeon display support" in the > kernel > > > > does not interfere with the ATI driver in a destructive way? > > > > > > I think, but im not sure, that the ati driver is loaded, if X was > > > started. But at this moment X was not started. > > > > Well, the kernel module part is probably loaded earlier. > > > > > > > 2. the fan starts spinning at the boot screen even if the radeon > display > > > > > driver was loaded (and the ati driver is not just loaded yet). > > > > > > > > I'm not sure what you mean exactly. Can you clarify, please?dann > > > > > > Well, i see it looks a little bit confused what i wrote. What i mean is, > > > that the fan at this time starts to rotate, when the "radeon display > > > support" is loaded at boot time. > > > > I see. > > > > Can you please check: > > - What happens if you don't use the ATI driver at all (please unload its > kernel > > module before loading the kernel's radeon one)? > > - Does the loading of the radeon module trigger the fan automatically? > > - What happens if the fan module is not loaded at that time? > > I think i got the problem: > I made some test. The problem is the module radeonfb. If i compile > radeonfb as a module there is no problem with the fan control. But if i > compile it into the kernel, than fan control will be destroyed and the > fan doesn't stop working from the beginning of the system start. > The other modules (fglrx, agp_gart) have no influence to the fan > behaviour. Thanks for having narrowed this, good job. Now, the question is why the radeonfb compiled in statically has this effect. Can you please attach your current .config to the bugzilla entry?
On Mon, 2007-10-15 at 22:32 +0200, Rafael J. Wysocki wrote: > > Thanks for having narrowed this, good job. > > Now, the question is why the radeonfb compiled in statically has this effect. > > Can you please attach your current .config to the bugzilla entry? I don't have the beginning of an idea of why radeonfb would cause that indeed. (And especially not cause it when compiled as a module). Ben.
On Monday, 15 October 2007 22:57, Benjamin Herrenschmidt wrote: > > On Mon, 2007-10-15 at 22:32 +0200, Rafael J. Wysocki wrote: > > > > Thanks for having narrowed this, good job. > > > > Now, the question is why the radeonfb compiled in statically has this > effect. > > > > Can you please attach your current .config to the bugzilla entry? > > I don't have the beginning of an idea of why radeonfb would cause that > indeed. (And especially not cause it when compiled as a module). Me too, but I suspect an initialization ordering issue (or something along these lines).