|
Bugzilla – Full Text Bug Listing |
| Summary: | YAST SL10 Installer should check for existing swap partitions | ||
|---|---|---|---|
| Product: | [openSUSE] SUSE LINUX 10.0 | Reporter: | Thomas Meindl <twm.mst> |
| Component: | Mobile Devices | Assignee: | Forgotten User ZhJd0F0L3x <forgotten_ZhJd0F0L3x> |
| Status: | VERIFIED FIXED | QA Contact: | Klaus Kämpf <kkaempf> |
| Severity: | Enhancement | ||
| Priority: | P5 - None | ||
| Version: | RC 1 | ||
| Target Milestone: | SUSE Linux 10.1 | ||
| Hardware: | i686 | ||
| OS: | All | ||
| Whiteboard: | |||
| Found By: | Other | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
|
Description
Thomas Meindl
2005-09-21 08:59:07 UTC
I'd call this an enhancement. Reassigned. So far we do not want to do partitioning suggestions that span more than one disk. This leads to problems as soon as people change disks. I would prefer to fix kpowersave that it can handle multiple swaps on multiple disks. It could for example look at /etc/fstab or /proc/swaps to find out which swap belongs to the currently running system. Well that's reasonable. Shall I close this bug as INVALID or WONTFIX and open a new one regarding powersave? No need to create a new bug. I already reassigned this bug to the maintainer of kpowersave. AFAIK the kernel itself gets confused if there is more than one swap partition. We could ugly workaround the problem by disabling all other swap partitions than the resume=/dev/xxxY partion, so that the kernel only uses that one to write data to. Seife, you know more about this stuff, close this one if changes are too complex... i'll have to ask Pavel on that. Pavel, what is the current status wrt multiple swap partitions? If it works at all, which partition will contain the swsusp signature? and which of them do i need to give as "resume=" device? I'll do some tests on this, but due to other high priority tasks these will have to wait. IIRC, kernel should now support swsusp with two swaps active (but will only use one of them). It is very little tested code path, though. Ok. I'll start testing it. I assume that the suspend header will be written to the partition given by "resume=..." on initial boot? Yes, it should work like that. i fixed this in powersave for 10.1: - we check if the swap partition given in "resume=" is available and only error out if it is not. - i tested multiple swap partitions - they work fine |