Re: Totally broken PCI PM calls

[Stefan Seyfried <seife@suse.de>]

Hello,

Pavel Machek wrote:

>>p.s. I find the /sys/power/disk file mildly cryptic, maybe
>>    other folk will find the attached patch slightly more
>>    informative about what this interface can do.
> 
>>--- 1.19/kernel/power/disk.c	Thu Sep  9 08:45:13 2004
>>+++ edited/kernel/power/disk.c	Fri Oct  1 11:01:41 2004
>>@@ -282,7 +282,14 @@
>> 
>> static ssize_t disk_show(struct subsystem * subsys, char * buf)
>> {
>>-	return sprintf(buf,"%s\n",pm_disk_modes[pm_disk_mode]);
>>+	return sprintf(buf,"%s%s %s%s %s%s\n",
>>+		(pm_disk_mode == pm_ops->pm_disk_mode) ? "*" : "",
>>+			pm_disk_modes[pm_ops->pm_disk_mode],
>>+		(pm_disk_mode == PM_DISK_SHUTDOWN) ? "*" : "",
>>+			pm_disk_modes[PM_DISK_SHUTDOWN],
>>+		(pm_disk_mode == PM_DISK_REBOOT) ? "*" : "",
>>+			pm_disk_modes[PM_DISK_REBOOT]
>>+		);
>> }
>
> Hmm, its interface change,

yes, this would break _my_ userspace app that checks for successfully
setting shutdown mode by reading out the file, but well, i could live
with that.

>  and was not /sys expected to be "one file, one value"?

# cat /sys/power/state
standby mem disk

so one could assume that /sys/power/disk would have similar semantics.
OTOH, /sys/power/state does not indicate the active state but
/sys/power/disk does, so they _are_ different.
-- 
Stefan Seyfried, QA / R&D Team Mobile Devices, SUSE LINUX AG Nürnberg.

"Any ideas, John?"
"Well, surrounding them's out."