Red Hat Bugzilla – Bug 113958
e1000 unreliable load after resume from ACPI sleep
Last modified: 2014-03-16 22:41:35 EDT
Using acpi sleep to suspend-to-ram, on resume, I attempt to load
the e1000 driver.
This fails with:
Jan 19 19:18:26 nostromo kernel: Intel(R) PRO/1000 Network Driver -
Jan 19 19:18:26 nostromo kernel: Copyright (c) 1999-2003 Intel
Jan 19 19:18:26 nostromo kernel: PCI: Enabling device 0000:02:01.0
(0000 -> 0003)
Jan 19 19:18:26 nostromo kernel: The EEPROM Checksum Is Not Valid
Jan 19 19:18:26 nostromo kernel: e1000: probe of 0000:02:01.0 failed
This behaves 'better' in 2.6.2-1.87, in that I only occasionally see
resume-from-suspend weirdness, instead of every time. (say, maybe one
of 10 times?)
Haven't seen this in a few weeks. Will reopen if it comes back.
Oh, wait. Haven't seen it because due to 4/4 I can't suspend and
Any different behavior with 2.6.5-1.326?
Haven't suspended enough to get a good handle. Will let you know in a
week or so. :/
Status is that there are enough other issues (airo driver *really*
doesn't like suspend, and doesn't unload correctly) that I haven't got
a chance to verify it. I can reopen if it comes back.