On Friday, March 14, 2008 8:44 am Rafael J. Wysocki wrote: > On Friday, 14 of March 2008, Matthew Wilcox wrote: > > On Fri, Mar 14, 2008 at 03:34:34PM +0100, Rafael J. Wysocki wrote: > > > On Friday, 14 of March 2008, Matthew Wilcox wrote: > > > > On Thu, Mar 13, 2008 at 09:44:08PM -0700, Greg KH wrote: > > > > > And why "_later"? What's wrong with the normal resume time? > > > > > > > > He said: > > > > > asus_hides_smbus_lpc_ich6() is called with interrupt disabled in > > > > > resume time, but ioremap_nocache() and iounmap() can't be called > > > > > with interrupt disabled. Below is my debug fix. If you have better > > > > > fix, please speak. > > > > > > > > Maybe _resume is run too early and could be run later when interrupts > > > > are enabled? > > > > > > That's what Shaohua is trying to achieve, AFAICS. > > > > Shaohua is adding an extra class of resume_later fixups and moving this > > one to it. I'm saying that maybe resume fixups should all be run later. > > Yes, I thought about that too. > > > But I haven't looked at the dependencies here. That may not be possible. > > Well, I'm not sure how to find out (reliably) who uses it. Seems a little risky, there are quite a few resume fixups (looks like at least 18 unique routines) but it might be ok as long as someone was willing to audit them all for irq enable safety... Jesse -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
