On Mon, Jul 03, 2006 at 11:46:32PM +0200, Pavel Machek wrote: > Interesting datapoint would be "does it work ok with 32-bit kernel"? > Some live-CD might help answer that question.
This hint is what it took to get everything going for me. I've figured out 32-bit s2ram for my laptop, and made some progress debugging 64-bit. I started by building a 32-bit s2ram and using it while running the 32-bit 2.6.17 kernel from knoppix 5.0.1. Here are the results: s2ram -f -a 1 s2ram -f -a 2 s2ram -f -a 3 machine functional after resume, no video s2ram -f -p -s s2ram -f -s Any attempt to use -s causes s2ram to segfault like this: [EMAIL PROTECTED] ./s2ram -f -s Switching from vt1 to vt1 Calling save_state Get video state buffer size failed Allocated buffer at 0x20000 (base is 0x0) ES: 0x2000 EBX: 0x0000 Segmentation fault I'll be happy to help debug it if someone wants to try. s2ram -f -p SUCCESS! POSTing the video bios after resume is what's needed for this laptop. On the console, nothing works quite right. After a suspend/resume with no framebuffer (booted with vga=0), The console is OK, but it's shrunk to a quarter-size rectangle in the middle of the screen. Before suspend, it was a normal vga console filling the whole screen. With framebuffer, the result is the same: a small vga console in the middle of the screen (no fb after resume). Only now, the contents of the console is just random green and pink character cells. The system is functional, though. Further investigation shows that it works within X, whether or not the boot console is using framebuffer. I'm not unloading any modules, or doing anything other than running "s2ram -f -p". After resume into X with a framebuffer console, the console VTs are messed up, but differently than if suspend/resume is done on an fb console. When doing suspend/resume in X, the konsole where s2ram is run gets about 50 phantom '\n' keypresses. The output of s2ram looks like this: [EMAIL PROTECTED] /tmp/s2ram -f -p Switching from vt5 to vt1 Calling do_post Calling INT 0x15 (F000:F859) EAX is 8C15F04 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 5F01 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 40005F00 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 40005F03 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 705F02 Leaving interrupt call. Calling INT 0x10 (C000:0D18) EAX is 704F14 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is F0005F15 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 705F15 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 705F15 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 25F15 Leaving interrupt call. Calling INT 0x15 (F000:F859) EAX is 805FFF Leaving interrupt call. switching back to vt5 [EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED] ... s2ram -f -a 1 -s not tried So all that leads me to this patch: diff -ur suspend-0.2.orig/whitelist.c suspend-0.2/whitelist.c --- suspend-0.2.orig/whitelist.c 2006-06-21 07:52:12.000000000 -0400 +++ suspend-0.2/whitelist.c 2006-07-04 14:00:53.000000000 -0400 @@ -54,6 +54,10 @@ { "Hewlett-Packard", "HP Compaq nx6125 *", "", "", VBE_SAVE|NOFB }, { "Hewlett-Packard", "HP Compaq nc6230 *", "", "", VBE_SAVE|NOFB }, { "Hewlett-Packard", "HP Compaq nx8220 *", "", "", VBE_SAVE|NOFB }, + /* The full match is: + * { "Hewlett-Packard ", "Pavilion zv5000 (DP523AV) ", "F.35", "F.35", VBE_POST|NOFB }, + * but I don't see any reason to match that tightly. - Jason Lunz */ + { "Hewlett-Packard*", "Pavilion zv5000 *", "", "", VBE_POST|NOFB }, { "Hewlett-Packard", "Presario R4100 *", "", "", S3_BIOS|S3_MODE }, /* R51 and T43 confirmed by Christian Zoz */ { "IBM", "1829*", "ThinkPad R51", "", 0 }, What's the policy on strict matching? I cleared out sys_version and bios_version since most other entries don't use them, and I use looser wildcard matches for sys_vendor and sys_product. Jason Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Suspend-devel mailing list Suspend-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/suspend-devel