~vcs-imports/suspend/trunk : changes

~vcs-imports/suspend/trunk » Changes from revision 455

From Revision 455 to 436

Rev	Summary	Authors	Date
455	Warning: this CVS no longer contains latest version of suspe... Warning: this CVS no longer contains latest version of suspend tools. Latest version is maintained in git://git.kernel.org/pub/scm/linux/kernel/git/rafael/suspend-utils.git	pavel	14 years ago
454	Kurt J. Bosch wrote: Kurt J. Bosch wrote: Fbsplash knows splash types 'suspend' and 'resume' for which special items can be defined in a theme. Use them instead of using 'bootup' allways which is wrong.	pavel	14 years ago
453	Replace VT_ACTIVATE ioctl call with userspace waiting in a l... Replace VT_ACTIVATE ioctl call with userspace waiting in a loop (a loop can timeout, VT_ACTIVATE can not and sometimes it really hangs for ever).	arekm	14 years ago
452	Patch from Luca Tettamanti : Patch from Luca Tettamanti : > > > One of our users discovered a bug in the suspend program which may cause > > > unplanned program termination. ... > > > Gentoo bug: > > > https://bugs.gentoo.org/show_bug.cgi?id=238511 > > > > > > ferret <ferret@explodingferret.com> writes: > > >> When I first configured and tried s2disk, it seemed to work > > > (splash > > >> progress bar and all) up to the point when I would expect it to > > > power > > >> off the system. It then prints an S and a newline and then > > > returns to > > >> the shell. > > >> > > >> I tracked the problem down to this C code in suspend.c and > > > resume.c: > > >> > > >> 1388: unsigned int size = sizeof(struct swsusp_header); > > >> 1389: unsigned int shift = (resume_offset + 1) * page_size - > > > size; > > >> > > >> Since I was using a swap file, my resume_offset is 528894, > > > page_size > > >> is 4096, and size is 28 (I have checked these values by > > > inserting a > > >> printf in the above code): > > The bug and the analisys are correct, but the fix only works for 64bit systems. ... > > lseek takes an off_t which is a long; the patch works fine on x86_64, > > but on plain x86 a long is still 32bit wide so the value passed to > > lseek is truncated. > > I believe that the proper fix is to use lseek64.	pavel	15 years ago
451	Add hp nx6315 to the whitelist. Add hp nx6315 to the whitelist.	pavel	15 years ago
450	Add 4 HP workstations, info provided by HP. Add 4 HP workstations, info provided by HP.	seife	15 years ago
449	Basic error checking to avoid a segfault when running as non... Basic error checking to avoid a segfault when running as non-root user (https://bugzilla.novell.com/show_bug.cgi?id=439337)	seife	15 years ago
448	whitelist update whitelist update	seife	15 years ago
447	s2ram whitelist autodiscovery from Alan Jenkins. s2ram whitelist autodiscovery from Alan Jenkins.	pavel	15 years ago
446	First round of whitelist updates. First round of whitelist updates.	seife	15 years ago
445	Allow the user to test the image saving and loading function... Allow the user to test the image saving and loading functionality of s2disk by making it read data from a regular file instead of the kernel. This allow one to debug the saving/loading code and benchmark it without causing the kernel to handle devices and create the image.	rjwysocki	15 years ago
444	Add configuration option "theads = " that makes s2disk/s2bot... Add configuration option "theads = " that makes s2disk/s2both use additional threads for saving the image, which may speed up the saving of compressed and encrypted images on multicore boxes. The idea is that the image pages read from the kernel are first placed in a work buffer, which is then compressed and its (compressed) contents are written into one of four "write" buffers (those buffers are used in a round robin manner). Then, another thread encrypts the contents of the work buffers and places the (encrypted) data into a "save" buffer that is read by the third thread which finally writes the data to disk. It turns out that with this patch applied one can speed up the saving of encrypted images quite significantly. For example, on my test box the speed of "threaded" saving of an encrytpded image is of the order of 60 MB/s, while for the same operation without threads the average speed is about 40 MB/s.	rjwysocki	15 years ago
443	From: Holger Macht <hmacht@suse.de>, Luca Tettamanti <kronos... From: Holger Macht <hmacht@suse.de>, Luca Tettamanti <kronos.it@gmail.com> The current DirectFB implementation handles keyboard events in a separate thread and this thread is frozen along with all other userspace during image writing. Because of this, keyboard presses are blocked when splashy is active which prevents aborting suspend or setting the reboot mode. Patch changes the keyboard handling and reads all keyboard events from /dev/input/by-path/*kbd instead of from stdin. This makes the keyboard handling independent from any splash implementation. Further information and references are here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=411061 https://bugzilla.novell.com/show_bug.cgi?id=293826	rjwysocki	15 years ago
442	Add Hewlett-Packard Presario F500 to the whitelist Add Hewlett-Packard Presario F500 to the whitelist	tdykstra	15 years ago
441	Fix HP BIOS borkage on 2510p Fix HP BIOS borkage on 2510p	seife	15 years ago
440	Whitelist Fujitsu Siemens Esprimo D9500 Whitelist Fujitsu Siemens Esprimo D9500	tdykstra	15 years ago
439	Whitelist update. Whitelist update.	seife	15 years ago
438	Big whitelist update. Big whitelist update.	seife	15 years ago
437	Allow the user to reboot and try again if the freezing of ta... Allow the user to reboot and try again if the freezing of tasks fails during resume from hibernation.	rjwysocki	15 years ago
436	Add: Acer AcerPower 2000, Aspire 5630, TravelMate 5320, BenQ... Add: Acer AcerPower 2000, Aspire 5630, TravelMate 5320, BenQ Joybook S32, HP Comaq 6820s, 8710p, Pavilion dv2000, Lenovo 3000C100 model 0761, 3000N200 mod.0769BBG, Medion MS-6714, MIM2080, SONY VGN-FE21B	seife	15 years ago

Older »