Bug 70601 - INTx doesn't work because core leaves PCI_COMMAND_INTX_DISABLE set
Summary: INTx doesn't work because core leaves PCI_COMMAND_INTX_DISABLE set
Status: RESOLVED CODE_FIX
Alias: None
Product: Drivers
Classification: Unclassified
Component: PCI (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Bjorn Helgaas
URL: https://lkml.kernel.org/r/20140114181...
Keywords:
Depends on:
Blocks:
 
Reported: 2014-02-14 20:37 UTC by Bjorn Helgaas
Modified: 2014-06-03 23:42 UTC (History)
0 users

See Also:
Kernel Version: 3.11
Subsystem:
Regression: No
Bisected commit-id:


Attachments
v3.11 dmesg showing problem (44.48 KB, text/plain)
2014-02-14 20:37 UTC, Bjorn Helgaas
Details
Enable INTx in PCI core (1.62 KB, patch)
2014-02-14 20:39 UTC, Bjorn Helgaas
Details | Diff
v3.11 dmesg with patch (problem fixed) (48.38 KB, text/plain)
2014-02-14 20:40 UTC, Bjorn Helgaas
Details
lspci -vv output (with patch) (806 bytes, text/plain)
2014-02-14 20:43 UTC, Bjorn Helgaas
Details

Description Bjorn Helgaas 2014-02-14 20:37:33 UTC
Created attachment 126171 [details]
v3.11 dmesg showing problem

On Intel Baytrail, with the BIOS "EHCI only" option selected, the BIOS leaves PCI_COMMAND_INTX_DISABLE set for the EHCI device.  The driver does nothing with that bit, so it never receives any interrupts.

This results in errors like these:

  usb 1-1: new high-speed USB device number 2 using ehci-pci
  usb 1-1: device not accepting address 2, error -110
  usb 1-1: new high-speed USB device number 3 using ehci-pci
  usb 1-1: device not accepting address 3, error -110
  ...
Comment 1 Bjorn Helgaas 2014-02-14 20:39:44 UTC
Created attachment 126181 [details]
Enable INTx in PCI core
Comment 2 Bjorn Helgaas 2014-02-14 20:40:20 UTC
Created attachment 126191 [details]
v3.11 dmesg with patch (problem fixed)
Comment 3 Bjorn Helgaas 2014-02-14 20:43:28 UTC
Created attachment 126201 [details]
lspci -vv output (with patch)

Note You need to log in before you can comment on or make changes to this bug.