On Sat, 2010-01-16 at 23:56 +0200, Igor M. Liplianin wrote: > On 16 января 2010 21:55:52 Andy Walls wrote: > > > cx25840 3-0044: IRQ Status: tsr rsr rto > > > cx25840 3-0044: IRQ Enables: rse rte roe > > > cx25840 3-0044: rx read: 9046778 ns mark > > > cx25840 3-0044: rx read: 2206333 ns space > > > cx25840 3-0044: rx read: 606926 ns mark > > > cx25840 3-0044: rx read: end of rx > > > cx25840 3-0044: IRQ Status: tsr rsr rto > > > cx25840 3-0044: IRQ Enables: rse rte roe > > > cx25840 3-0044: rx read: 9055815 ns mark > > > cx25840 3-0044: rx read: 2203519 ns space > > > cx25840 3-0044: rx read: 582481 ns mark > > > cx25840 3-0044: rx read: end of rx > > > > This is still good. :) > > > > Those are NEC repeat sequences, but you probably know that already. > Remote works until that point > > > > > > irq 16: nobody cared (try booting with the "irqpoll" option) > > > Pid: 2971, comm: X Not tainted 2.6.33-rc4 #3 > > > Call Trace: > > > [<c1054700>] ? __report_bad_irq+0x24/0x69 > > > > [...] > > > > > [<c1425105>] ? syscall_call+0x7/0xb > > > handlers: > > > [<c1332132>] (usb_hcd_irq+0x0/0x59) > > > [<f8aafd88>] (cx23885_irq+0x0/0x4e0 [cx23885]) > > > > I have checked in more changes to > > > > http://linuxtv.org/hg/~awalls/cx23885-ir2 > > > > Please test again using these module parameters: > > > > modprobe cx25840 ir_debug=2 debug=2 > > modprobe cx23885 ir_input_debug=2 irq_debug=7 debug=7 > > > > > > I am looking for logging of the interrupt statuses and enables. They > > should look something like this: > > > > > > kernel: cx23885[0]/0: pci_status: 0x08000000 pci_mask: 0x08000001 > > [...] > > kernel: cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > > kernel: cx25840 6-0044: AV Core IRQ status (entry): ir > > kernel: cx25840 6-0044: AV Core ir IRQ status: 0x31 disables: 0x20 > > kernel: cx25840 6-0044: IR IRQ Status: tsr rsr rto > > kernel: cx25840 6-0044: IR IRQ Enables: rse rte roe > > kernel: cx25840 6-0044: AV Core audio IRQ status: 0x80 disables: 0xff > > kernel: cx25840 6-0044: AV Core audio MC IRQ status: 0x2000 enables: > > 0x0000 kernel: cx25840 6-0044: AV Core video IRQ status: 0x01a7 disables: > > 0xffff kernel: cx25840 6-0044: AV Core IRQ status (exit): > > > > > > But I was able to reproduce something like this when changing enable the > > TSR interrupt enables using v4l2-dbg to change the register manually: > > > > kernel: cx23885[0]/0: pci_status: 0x08300000 pci_mask: 0x08000001 > > [...] > > kernel: cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > > kernel: cx25840 6-0044: AV Core IRQ status (entry): > > <---- no irq flags (all 0's) kernel: cx25840 6-0044: AV Core ir IRQ > > status: 0x00 disables: 0x00 <---- all 0's kernel: cx25840 > > 6-0044: AV Core audio IRQ status: 0x00 disables: 0x00 <---- all 0's > > kernel: cx25840 6-0044: AV Core audio MC IRQ status: 0x0000 enables: 0x0000 > > <---- all 0's kernel: cx25840 6-0044: AV Core video IRQ status: 0x0000 > > disables: 0x0000 <---- all 0's kernel: cx25840 6-0044: AV Core IRQ > > status (exit): > > > > So there are some conditions where the AV Core can signal an interrupt, > > but not be ready to be read over the I2C bus. > > > > I have added code to count when these happen and handle them as spurious > > interrupts. However, if the code gets too many (20) consecutive > > spurious interrupts without at least one real one, it disables the AV > > Core interrupt. > > > > > > Regards, > > Andy > Spurious interrupts counter reaches #20 almost immediately after modprobe. > There is no time to press a key. > > DVB: registering adapter 0 frontend 0 (Montage Technology DS3000/TS2020)... > cx23885_dev_checkrevision() Hardware revision = 0xb0 > cx23885[0]/0: found at 0000:01:00.0, rev: 2, irq: 16, latency: 0, mmio: 0xfe800000 > cx23885 0000:01:00.0: setting latency timer to 64 > IRQ 16/cx23885[0]: IRQF_DISABLED is not guaranteed on shared IRQs > cx23885[0]/0: pci_status: 0x083f4000 pci_mask: 0x08000000 > cx23885[0]/0: vida_status: 0x00000000 vida_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts1_status: 0x00000000 ts1_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts2_status: 0x00000000 ts2_mask: 0x00000000 count: 0xc7381f2a > cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > cx25840 3-0044: AV Core IRQ status (entry): > cx25840 3-0044: AV Core ir IRQ status: 0x00 disables: 0x00 > cx25840 3-0044: AV Core audio IRQ status: 0x00 disables: 0x00 > cx25840 3-0044: AV Core audio MC IRQ status: 0x0000 enables: 0x0000 > cx25840 3-0044: AV Core video IRQ status: 0x0000 disables: 0x0000 > cx25840 3-0044: AV Core IRQ status (exit): > cx23885[0]/0: consecutive spurious AV Core interrupt #1 > [...] > cx23885[0]/0: pci_status: 0x08004000 pci_mask: 0x08000000 > cx23885[0]/0: vida_status: 0x00000000 vida_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts1_status: 0x00000000 ts1_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts2_status: 0x00000000 ts2_mask: 0x00000000 count: 0xc7381f2a > cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > cx25840 3-0044: AV Core IRQ status (entry): > cx25840 3-0044: AV Core ir IRQ status: 0x00 disables: 0x00 > cx25840 3-0044: AV Core audio IRQ status: 0x00 disables: 0x00 > cx25840 3-0044: AV Core audio MC IRQ status: 0x0000 enables: 0x0000 > cx25840 3-0044: AV Core video IRQ status: 0x0000 disables: 0x0000 > cx25840 3-0044: AV Core IRQ status (exit): > cx23885[0]/0: consecutive spurious AV Core interrupt #20 > cx23885[0]: disabling AV Core/IR interrupt: too many spurious interrupts > input: cx23885 IR (TeVii S470) as /class/input/input24 > Creating IR device irrcv0 OK. I think I finally have guessed what is going on: 1. The cx23885_irq_handler is called for the AV_CORE when something else in the cx23885 or cx25840 module is accessing that I2C bus. 2. The I2C bus adapter mutex in the i2c_subsystem stays locked so the cx23885_irq_handler() and cx25840_irq_handler() cannot read the AV core registers since the I2C subsystem returns -EINVAL and 0 for the data. 3. The interrupt handler keeps getting called because it never clears the interrupt condition in the AV Core. I think I have to do some work in the cx25840 module and the cx23885 module to handle locking of the AV Core I2C client and I2C bus 3. > I commented out interrupt disabling line in order to get some debug to you. > But it occupied gigabytes on hard disk quickly. > > [...] > cx23885[0]/0: consecutive spurious AV Core interrupt #95 > cx23885[0]: disabling AV Core/IR interrupt: too many spurious interrupts > cx23885[0]/0: pci_status: 0x08304000 pci_mask: 0x08000000 > cx23885[0]/0: vida_status: 0x00000000 vida_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts1_status: 0x00000000 ts1_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts2_status: 0x00000000 ts2_mask: 0x00000000 count: 0xc7381f2a > cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > cx25840 3-0044: AV Core IRQ status (entry): ir > cx25840 3-0044: AV Core ir IRQ status: 0x31 disables: 0x20 > cx25840 3-0044: IR IRQ Status: tsr rsr rto > cx25840 3-0044: IR IRQ Enables: rse rte roe > cx25840 3-0044: AV Core audio IRQ status: 0x80 disables: 0xff > cx25840 3-0044: AV Core audio MC IRQ status: 0x2000 enables: 0x0000 > cx25840 3-0044: AV Core video IRQ status: 0x01a7 disables: 0xffff > cx25840 3-0044: AV Core IRQ status (exit): > cx23885[0]/0: pci_status: 0x08004000 pci_mask: 0x08000000 > cx23885[0]/0: vida_status: 0x00000000 vida_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts1_status: 0x00000000 ts1_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts2_status: 0x00000000 ts2_mask: 0x00000000 count: 0xc7381f2a > cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > cx25840 3-0044: AV Core IRQ status (entry): > cx25840 3-0044: AV Core ir IRQ status: 0x20 disables: 0x20 > cx25840 3-0044: AV Core audio IRQ status: 0x80 disables: 0xff > cx25840 3-0044: AV Core audio MC IRQ status: 0x2000 enables: 0x0000 > cx25840 3-0044: AV Core video IRQ status: 0x01a7 disables: 0xffff > cx25840 3-0044: AV Core IRQ status (exit): > cx23885[0]/0: consecutive spurious AV Core interrupt #1 > cx25840 3-0044: rx read: 9061000 ns mark > cx25840 3-0044: rx read: 2207667 ns space > cx25840 3-0044: rx read: 586333 ns mark > cx25840 3-0044: rx read: end of rx > cx23885 0000:01:00.0: NEC IR auto repeat detected > cx23885[0]/0: pci_status: 0x08304000 pci_mask: 0x08000000 > cx23885[0]/0: vida_status: 0x00000000 vida_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts1_status: 0x00000000 ts1_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts2_status: 0x00000000 ts2_mask: 0x00000000 count: 0xc7381f2a > cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > cx25840 3-0044: AV Core IRQ status (entry): > cx25840 3-0044: AV Core ir IRQ status: 0x20 disables: 0x20 > cx25840 3-0044: AV Core audio IRQ status: 0x80 disables: 0xff > cx25840 3-0044: AV Core audio MC IRQ status: 0x2000 enables: 0x0000 > cx25840 3-0044: AV Core video IRQ status: 0x01a7 disables: 0xffff > cx25840 3-0044: AV Core IRQ status (exit): > cx23885[0]/0: consecutive spurious AV Core interrupt #2 > cx23885[0]/0: pci_status: 0x08004000 pci_mask: 0x08000000 > cx23885[0]/0: vida_status: 0x00000000 vida_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts1_status: 0x00000000 ts1_mask: 0x00000000 count: 0x0 > cx23885[0]/0: ts2_status: 0x00000000 ts2_mask: 0x00000000 count: 0xc7381f2a > cx23885[0]/0: (PCI_MSK_AV_CORE 0x08000000) > [...] I'll have to think aboput these.... I'll fix the I2C bus problem first and then come back to this. Thanks for testing. Regards, Andy -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
