On Sun, 17 Feb 2019, Alexandre Oliva wrote: > That's a reasonable guess, but I don't think so. I do have PATA_AMD > enabled as a module indeed, but it's not even loaded, much as I can > tell, whereas PATA_CS5536 is built into the kernel image, and dmesg > says: > > [ 4.460000] scsi host0: pata_cs5536 > [ 4.464000] scsi host1: pata_cs5536 > [ 4.464000] ata1: PATA max UDMA/100 cmd 0x1f0 ctl 0x3f6 bmdma 0x4c60 irq 14 > [ 4.464000] ata2: DUMMY > [ 4.464000] pcnet32: [...] > [ 4.644000] random: [...] > [ 5.908000] irq 14: nobody cared (try booting with the "irqpoll" option)
Is there an MMIO completion barrier missing there somewhere by any chance causing an IRQ that has been handled already to be redelivered because an MMIO write meant to clear the IRQ at its origin at handler's completion has not reached its destination before interrupts have been reenabled in the issuing CPU? Just a thought. Maciej
