Bug 5060
| Summary: | acpi_video_device_write_state() busted | ||
|---|---|---|---|
| Product: | ACPI | Reporter: | Ariel T. Glenn (ariel) |
| Component: | Power-Video | Assignee: | Luming Yu (luming.yu) |
| Status: | CLOSED CODE_FIX | ||
| Severity: | normal | CC: | acpi-bugzilla |
| Priority: | P2 | ||
| Hardware: | i386 | ||
| OS: | Linux | ||
| Kernel Version: | 2.6.13-rc6 | Subsystem: | |
| Regression: | --- | Bisected commit-id: | |
| Attachments: | patch against 2.6.13-rc6 | ||
Created attachment 5720 [details]
patch against 2.6.13-rc6
Thanks for the analysis, this patch should fix your problem.
I've been running with that very fix (with a different variable name) and it works fine. Thanks. Ariel closed means in base kernel. So just mark it as resolved. applied Shipped in 2.6.13-git9. Closing. |
Distribution:FC4, vanilla kernel NOT from FC4. Hardware Environment: Dell XPS Gen 2 laptop Software Environment: most modules unloaded, running from console, most services shut down Problem Description: acpi_video_device_write_state() calls acpi_video_device_set_state() which calls acpi_evaluate_integer() with fourth arg NULL, causing an error so the set state call bails. Steps to reproduce: echo 0x20001f > /proc/acpi/debug_level echo 0x80000001 > /proc/acpi/video/VID/LCD/state check in log; you'll see something like this: Aug 13 18:31:03 localhost kernel: acpi_video-0782 [13] acpi_video_device_writ: ----Entry Aug 13 18:31:03 localhost kernel: acpi_video-0302 [14] acpi_video_device_set_: ----Entry Aug 13 18:31:03 localhost kernel: acpi_utils-0250 [15] acpi_evaluate_integer : ----Entry Aug 13 18:31:03 localhost kernel: acpi_utils-0253 [15] acpi_evaluate_integer : ----Exit- ****Exception****: AE_BAD_PARAMETER Aug 13 18:31:03 localhost kernel: acpi_video-0307 [14] acpi_video_device_set_: ----Exit- 0000000000001001 Aug 13 18:31:03 localhost kernel: acpi_video-0797 [13] acpi_video_device_writ: ----Exit- FFFFFFFFFFFFFFF2 Here's the call that's bad: status = acpi_evaluate_integer(device->handle, "_DSS", &args, NULL); (around line 305 in video.c) Here's why it's bad: acpi_status acpi_evaluate_integer ( acpi_handle handle, acpi_string pathname, struct acpi_object_list *arguments, unsigned long *data) ... ACPI_FUNCTION_TRACE("acpi_evaluate_integer"); if (!data) return_ACPI_STATUS(AE_BAD_PARAMETER); (around line 240 in util.c) Since I don't know whether there should be an actual argument passed or whether acpi_evaluate_integer shouldn't do that check, I didn't attach a patch. I checked in http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/patches/release/2.6.13/broken-out/ but didn't find a fix for this. I'm not adding attachments unless someone thinks they really need them.