Le 14/08/2015 22:17, Bjorn Helgaas a écrit : > For 1), I think it is probably overkill to penalize all JMicron > devices. There's no reason to think NICs and SD/MMC/etc controllers > have the same issue. Or if there *is* reason to think that, you > should give evidence for it. if you don't want to penalize all JMicron models then a workaround would be to create a new feature : the ability for the user to disable async for a specific device, for example a kernel parameter ( no-async-for-jmicron ), or a more configurable mechanism, like a configuration file where the user can add the name of the device, a kind of blacklist in order to disable async only for a list of devices, it's only a workaround, the real solution would to be find the real root cause ( a bug in async method ? a design flaw in JMicron hardware ? a bios/acpi bug ? ) -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
