Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
On Don, 2012-03-22 at 00:05 +0100, Stephen Kitt wrote: On Wed, Sep 14, 2011 at 07:14:12PM +0200, dayer wrote: El mié, 14-09-2011 a las 11:23 +0200, Michel Dänzer escribió: Since several IRQs are affected, I suspect this is a fundamental problem in the kernel/BIOS rather than in any driver. I think the problem is my graphic card ATI Gigabyte HD4650 AGP 1GB. I'm using it one or two years without problem until April 2011, when IRQs problems started. In Windows XP there's no problem. And if I try connecting an old ATI 9600 AGP 256MB in Linux there's no problem with IRQs. I've been having a similar problem for a while now, with a similar graphics card (an XFX HD4650 AGP with 1GB of RAM): at every boot, regardless of irqpoll, the card's IRQ would end up being disabled. I hadn't bothered filing a bug because like Michel, I thought it was probably due to a BIOS bug. Today though I noticed a pattern: in every single instance, just after the IRQ got disabled, the kernel logged [ 55.404002] Disabling IRQ #16 [ 56.912023] hda-intel: azx_get_response timeout, switching to polling mode: last cmd=0x000f0001 So I wondered whether this might be somehow related to the snd-hda-intel driver attempting to initialize the HDMI audio output... I blacklisted the snd-hda-intel driver, and hey presto, no more errors. Does enabling HDMI audio with radeon.audio=1 prevent the problem as well? -- Earthling Michel Dänzer | http://www.amd.com Libre software enthusiast | Debian, X and DRI developer -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
Hi Michel, On Thu, Mar 22, 2012 at 08:08:57AM +0100, Michel Dänzer wrote: On Don, 2012-03-22 at 00:05 +0100, Stephen Kitt wrote: So I wondered whether this might be somehow related to the snd-hda-intel driver attempting to initialize the HDMI audio output... I blacklisted the snd-hda-intel driver, and hey presto, no more errors. Does enabling HDMI audio with radeon.audio=1 prevent the problem as well? Unfortunately not: [...] [0.00] Kernel command line: BOOT_IMAGE=/vmlinuz-3.2.0-2-686-pae root=/dev/mapper/vg--large--mirror-root ro quiet splash radeon.audio=1 [...] [ 52.612699] irq 16: nobody cared (try booting with the irqpoll option) [ 52.612709] Pid: 255, comm: plymouthd Not tainted 3.2.0-2-686-pae #1 [ 52.612713] Call Trace: [ 52.612730] [c1078a21] ? __report_bad_irq+0x1c/0x8d [ 52.612736] [c1078c2b] ? note_interrupt+0x122/0x18f [ 52.612742] [c10775cb] ? handle_irq_event_percpu+0x142/0x158 [ 52.612748] [c10791de] ? handle_level_irq+0x56/0x56 [ 52.612753] [c1077602] ? handle_irq_event+0x21/0x37 [ 52.612758] [c10791de] ? handle_level_irq+0x56/0x56 [ 52.612764] [c107923e] ? handle_fasteoi_irq+0x60/0x85 [ 52.612767] IRQ [c100cdcf] ? do_IRQ+0x2e/0x76 [ 52.612781] [c12c57f0] ? common_interrupt+0x30/0x38 [ 52.612787] [c12c007b] ? mutex_unlock+0x10/0x12 [ 52.612792] [c12c] ? __wait_on_bit+0x4c/0x54 [ 52.612795] handlers: [ 52.612844] [f82722fd] usb_hcd_irq [ 52.612860] [f82722fd] usb_hcd_irq [ 52.612893] [f83d9dc9] radeon_driver_irq_handler_kms [ 52.612898] Disabling IRQ #16 [ 54.112026] hda-intel: azx_get_response timeout, switching to polling mode: last cmd=0x000f0001 [ 54.121170] HDMI status: Codec=0 Pin=3 Presence_Detect=1 ELD_Valid=1 [ 54.121422] input: HDA ATI HDMI HDMI/DP,pcm=3 as /devices/pci:00/:00:01.0/:01:00.1/sound/card5/input8 I don't know whether it's relevant, but the card I'm using doesn't advertise HDMI output in its documentation; it has two DVI outputs, I don't have a DVI-HDMI adapter to try with it, so I don't know if there is actually any way of getting audio out of it. Regards, Stephen -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
Hi, On Wed, Sep 14, 2011 at 07:14:12PM +0200, dayer wrote: El mié, 14-09-2011 a las 11:23 +0200, Michel Dänzer escribió: Since several IRQs are affected, I suspect this is a fundamental problem in the kernel/BIOS rather than in any driver. I think the problem is my graphic card ATI Gigabyte HD4650 AGP 1GB. I'm using it one or two years without problem until April 2011, when IRQs problems started. In Windows XP there's no problem. And if I try connecting an old ATI 9600 AGP 256MB in Linux there's no problem with IRQs. I've been having a similar problem for a while now, with a similar graphics card (an XFX HD4650 AGP with 1GB of RAM): at every boot, regardless of irqpoll, the card's IRQ would end up being disabled. I hadn't bothered filing a bug because like Michel, I thought it was probably due to a BIOS bug. Today though I noticed a pattern: in every single instance, just after the IRQ got disabled, the kernel logged [ 55.404002] Disabling IRQ #16 [ 56.912023] hda-intel: azx_get_response timeout, switching to polling mode: last cmd=0x000f0001 So I wondered whether this might be somehow related to the snd-hda-intel driver attempting to initialize the HDMI audio output... I blacklisted the snd-hda-intel driver, and hey presto, no more errors. In the original submitter's case, both IRQ #16 and IRQ #17 got disabled; but on my system IRQ #16 is the VGA controller, and #17 the audio device. I noted too that on the first report mentioning IRQ #17, the handlers include azx_interrupt which belongs to snd-hda-intel... So it might be worth trying to add blacklist snd-hda-intel to /etc/modprobe.d/local.conf for example, and running update-initramfs -u to make sure it's not loaded from the initramfs. Regards, Stephen -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
On Mit, 2011-09-14 at 10:33 +0200, dayer wrote: Package: xserver-xorg-video-radeon Version: 1:6.14.2-1 Followup-For: Bug #628080 With irqpoll at boot the problem is still being. dmesg prints this: [ 1077.872450] irq 16: nobody cared (try booting with the irqpoll option) Do these messages also appear if you boot with irqpoll? Are you sure it takes effect? [ 1077.872667] handlers: [ 1077.872723] [a02895c9] radeon_driver_irq_handler_kms [ 1077.872729] Disabling IRQ #16 [ 1300.396474] irq 17: nobody cared (try booting with the irqpoll option) [...] [ 1300.396536] handlers: [ 1300.396559] [a0004a84] skge_intr [ 1300.396561] Disabling IRQ #17 Since several IRQs are affected, I suspect this is a fundamental problem in the kernel/BIOS rather than in any driver. -- Earthling Michel Dänzer | http://www.amd.com Libre software enthusiast | Debian, X and DRI developer -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
El mié, 14-09-2011 a las 11:23 +0200, Michel Dänzer escribió: On Mit, 2011-09-14 at 10:33 +0200, dayer wrote: Do these messages also appear if you boot with irqpoll? Are you sure it takes effect? Yes, I put irqpoll at grub's boot but I can't understand why irqpoll doesn't work. Yesterday and today I'm trying with options radeon modeset=0 in /etc/modprobe.d/radeon-kms.conf, and although the performance is odd according to the program (Chromium goes smoother than Iceweasel), the direct rendering is on and there is only a problem with an IRQ: dmesg: -- [13075.848407] irq 16: nobody cared (try booting with the irqpoll option) [13075.848421] Pid: 0, comm: swapper Tainted: G O 3.0.0-1-amd64 #1 [13075.848426] Call Trace: [13075.848430] IRQ [81093aba] ? __report_bad_irq+0x2c/0xb7 [13075.848455] [81093e93] ? note_interrupt+0x174/0x1f9 [13075.848464] [81092694] ? handle_irq_event_percpu+0x166/0x184 [13075.848473] [81267baa] ? acpi_pm_read+0x5/0x8 [13075.848482] [81066feb] ? timekeeping_get_ns+0xd/0x2a [13075.848490] [810926e6] ? handle_irq_event+0x34/0x52 [13075.848498] [8106cd08] ? arch_local_irq_save+0x14/0x1d [13075.848506] [8109459a] ? handle_fasteoi_irq+0x79/0x9b [13075.848514] [8100a9aa] ? handle_irq+0x1a/0x20 [13075.848521] [8100a6db] ? do_IRQ+0x45/0x9e [13075.848529] [81336e13] ? common_interrupt+0x13/0x13 [13075.848534] EOI [81026730] ? native_safe_halt+0x2/0x3 [13075.848549] [8100f26f] ? default_idle+0x49/0x83 [13075.848556] [8100827e] ? cpu_idle+0xa2/0xc4 [13075.848565] [8169fb65] ? start_kernel+0x3bd/0x3c8 [13075.848573] [8169f3c6] ? x86_64_start_kernel+0x102/0x10f [13075.848578] handlers: [13075.848624] [a0278906] radeon_driver_irq_handler [13075.848630] Disabling IRQ #16 Some lines have changed and they're different if KMS is on, but according to /proc/interrups IRQ #16 is also from radeon. Since several IRQs are affected, I suspect this is a fundamental problem in the kernel/BIOS rather than in any driver. I think the problem is my graphic card ATI Gigabyte HD4650 AGP 1GB. I'm using it one or two years without problem until April 2011, when IRQs problems started. In Windows XP there's no problem. And if I try connecting an old ATI 9600 AGP 256MB in Linux there's no problem with IRQs. -- http://elhistorias.com -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
On Sam, 2011-06-04 at 17:34 +0200, dayer wrote: El mié, 01-06-2011 a las 09:14 +0200, Michel Dänzer escribió: On Fre, 2011-05-27 at 01:30 +0200, dayer wrote: When the problem occurs, do the radeon numbers in /proc/interrupts increase? Yes. /proc/interrupts contains: CPU0 0: 44 IO-APIC-edge timer 1: 4 IO-APIC-edge i8042 6: 3 IO-APIC-edge floppy 7: 1 IO-APIC-edge parport0 8: 0 IO-APIC-edge rtc0 9: 0 IO-APIC-fasteoi acpi 12: 6 IO-APIC-edge i8042 14: 13995 IO-APIC-edge pata_via 15: 0 IO-APIC-edge pata_via 16: 41 IO-APIC-fasteoi radeon 17: 556661064 IO-APIC-fasteoi skge@pci::00:0a.0, hda_intel [...] The radeon numbers are increasing and the end this causes these lines in dmesg (and the system begins to run slow): Do the radeon numbers stop increasing after the messages about IRQ 16? [ 1679.754669] irq 16: nobody cared (try booting with the irqpoll option) [...] [ 1679.754896] Disabling IRQ #16 And another messages appear later also in dmesg: [ 4984.745387] irq 17: nobody cared (try booting with the irqpoll option) [...] [ 4984.745479] Disabling IRQ #17 I'm thinking about try booting with the irqpoll. What do you think? Probably can't hurt to try. Note that as IRQ 17 is also affected, this may not be a radeon specific problem at all but a more general IRQ problem on your system. Does booting with radeon.agpmode=-1 on the kernel command line help? Yes, it helps! The performance doesn't fall already, although... With it in /var/log/Xorg.0.log appears a new warning in addition to previous: (WW) RADEON(0): Direct rendering disabled Something's going wrong with that option, resulting in the DRI being disabled completely. As you've experienced with empathy, this is not desirable for various reasons. -- Earthling Michel Dänzer |http://www.vmware.com Libre software enthusiast | Debian, X and DRI developer -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
El vie, 17-06-2011 a las 11:15 +0200, Michel Dänzer escribió: On Sam, 2011-06-04 at 17:34 +0200, dayer wrote: Do the radeon numbers stop increasing after the messages about IRQ 16? Yes, it does. After the messages about IRQ 16 radeon numbers stops, and after the messages about IRQ 17 this line also stops increasing. Probably can't hurt to try. Note that as IRQ 17 is also affected, this may not be a radeon specific problem at all but a more general IRQ problem on your system. I'am testing it now with irqpoll and without radeon agpmode=-1, and there's no problem. Irqpoll looks the solution in my case. Something's going wrong with that option, resulting in the DRI being disabled completely. As you've experienced with empathy, this is not desirable for various reasons. -- http://elhistorias.com -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
El mié, 01-06-2011 a las 09:14 +0200, Michel Dänzer escribió: On Fre, 2011-05-27 at 01:30 +0200, dayer wrote: Are you using an OpenGL compositing manager such as gnome-shell or compiz? Now I'm using Metacity. However I've had the same problem with Mutter. When the problem occurs, do the radeon numbers in /proc/interrupts increase? Yes. /proc/interrupts contains: CPU0 0: 44 IO-APIC-edge timer 1: 4 IO-APIC-edge i8042 6: 3 IO-APIC-edge floppy 7: 1 IO-APIC-edge parport0 8: 0 IO-APIC-edge rtc0 9: 0 IO-APIC-fasteoi acpi 12: 6 IO-APIC-edge i8042 14: 13995 IO-APIC-edge pata_via 15: 0 IO-APIC-edge pata_via 16: 41 IO-APIC-fasteoi radeon 17: 556661064 IO-APIC-fasteoi skge@pci::00:0a.0, hda_intel 20: 33647 IO-APIC-fasteoi sata_via 21: 32145 IO-APIC-fasteoi ehci_hcd:usb1, uhci_hcd:usb2, uhci_hcd:usb3, uhci_hcd:usb4, uhci_hcd:usb5 22: 63 IO-APIC-fasteoi VIA8237 NMI: 40 Non-maskable interrupts LOC:1981218 Local timer interrupts SPU: 0 Spurious interrupts PMI: 40 Performance monitoring interrupts IWI: 0 IRQ work interrupts RES: 0 Rescheduling interrupts CAL: 0 Function call interrupts TLB: 0 TLB shootdowns TRM: 0 Thermal event interrupts THR: 0 Threshold APIC interrupts MCE: 0 Machine check exceptions MCP: 16 Machine check polls ERR: 1 MIS: 0 The radeon numbers are increasing and the end this causes these lines in dmesg (and the system begins to run slow): [ 1679.754669] irq 16: nobody cared (try booting with the irqpoll option) [ 1679.754682] Pid: 0, comm: swapper Tainted: G O 2.6.39-1-amd64 #1 [ 1679.754688] Call Trace: [ 1679.754692] IRQ [81092520] ? __report_bad_irq+0x3a/0xa2 [ 1679.754715] [810926f0] ? note_interrupt+0x168/0x1da [ 1679.754724] [81090e20] ? handle_irq_event_percpu +0x171/0x18f [ 1679.754733] [8104b547] ? __do_softirq+0x13b/0x19e [ 1679.754743] [8106618b] ? timekeeping_get_ns+0xe/0x2e [ 1679.754750] [81090e72] ? handle_irq_event+0x34/0x52 [ 1679.754759] [81063ffb] ? sched_clock_idle_wakeup_event +0x12/0x1c [ 1679.754767] [81092e3e] ? handle_fasteoi_irq+0x82/0xa4 [ 1679.754775] [8100aadb] ? handle_irq+0x1a/0x23 [ 1679.754781] [8100a384] ? do_IRQ+0x45/0xaa [ 1679.754790] [81332ad3] ? common_interrupt+0x13/0x13 [ 1679.754794] EOI [81026800] ? native_safe_halt+0x2/0x3 [ 1679.754810] [8100f70d] ? default_idle+0x4b/0x85 [ 1679.754817] [81008dd2] ? cpu_idle+0xab/0xe1 [ 1679.754826] [8169fc60] ? start_kernel+0x3e0/0x3eb [ 1679.754835] [8169f3c8] ? x86_64_start_kernel+0x102/0x10f [ 1679.754839] handlers: [ 1679.754843] [a0370ee0] (radeon_driver_irq_handler_kms +0x0/0x1c [radeon]) [ 1679.754896] Disabling IRQ #16 [ 1679.754896] Disabling IRQ #16 And another messages appear later also in dmesg: [ 4984.745387] irq 17: nobody cared (try booting with the irqpoll option) [ 4984.745396] Pid: 2358, comm: firefox-bin Tainted: G O 2.6.39-1-amd64 #1 [ 4984.745398] Call Trace: [ 4984.745401] IRQ [81092520] ? __report_bad_irq+0x3a/0xa2 [ 4984.745414] [810926f0] ? note_interrupt+0x168/0x1da [ 4984.745419] [81090e20] ? handle_irq_event_percpu +0x171/0x18f [ 4984.745423] [8104b547] ? __do_softirq+0x13b/0x19e [ 4984.745426] [81090e72] ? handle_irq_event+0x34/0x52 [ 4984.745430] [81092e3e] ? handle_fasteoi_irq+0x82/0xa4 [ 4984.745433] [8100aadb] ? handle_irq+0x1a/0x23 [ 4984.745436] [8100a384] ? do_IRQ+0x45/0xaa [ 4984.745440] [81332ad3] ? common_interrupt+0x13/0x13 [ 4984.745442] EOI [81338c12] ? system_call_fastpath +0x16/0x1b [ 4984.745447] handlers: [ 4984.745448] [a001c289] (skge_intr+0x0/0x3e9 [skge]) [ 4984.745474] [a032f34e] (azx_interrupt+0x0/0x13c [snd_hda_intel]) [ 4984.745479] Disabling IRQ #17 I'm thinking about try booting with the irqpoll. What do you think? It's a long shot, but maybe you could try if a 2.6.39 kernel helps. It's the same. Does booting with radeon.agpmode=-1 on the kernel command line help? Yes, it helps! The performance doesn't fall already, although... With it in /var/log/Xorg.0.log appears a new warning in addition to previous: (WW) RADEON(0): Direct rendering disabled And I can't run empathy due to: failed to create drawable Unable to select the newly created GLX context Run 'empathy --help' to see a full list of available command line options. (empathy:17688): empathy-WARNING **: Error in empathy init: Unable to select the newly created GLX context But the solution is better than the previous situation. --
Bug#628080: xserver-xorg-video-radeon: Low performance after a period of inactivity
On Fre, 2011-05-27 at 01:30 +0200, dayer wrote: Package: xserver-xorg-video-radeon Version: 1:6.14.1-1+b1 Severity: normal Hi, I haven't any problem with perfomance unless I leave the computer alone around two hours. If I don't use it in 20 minuts, when I back there's no problem. But with more time, e.g. 2 hours, when I back the performance is very low. In these situations I can't write fluently neither Iceweasel nor Gnome Terminal, for instance, and windows are minimized/restored very slowly, with cuts. If I restart GDM the problem is still and the solution is reboot the system. Are you using an OpenGL compositing manager such as gnome-shell or compiz? When the problem occurs, do the radeon numbers in /proc/interrupts increase? Kernel version (/proc/version): --- Linux version 2.6.38-2-amd64 (Debian 2.6.38-5) (b...@decadent.org.uk) (gcc version 4.4.6 (Debian 4.4.6-3) ) #1 SMP Sun May 8 13:51:57 UTC 2011 It's a long shot, but maybe you could try if a 2.6.39 kernel helps. [19.935] (II) RADEON(0): AGP card detected Does booting with radeon.agpmode=-1 on the kernel command line help? -- Earthling Michel Dänzer |http://www.vmware.com Libre software enthusiast | Debian, X and DRI developer -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org