> > I used 'git apply --check ' first, and got no errors, so > > I applied it, built, installed, and rebooted. > > could apply with > cat revert_*.;patch | patch -p1 Yinghai, When I got home from work tonight, I decided to make sure that I applied your patch correctly. I considered trying it once with cat/patch and once with 'git apply', and then comparing the results with 'diff'. I ended up becoming convinced that 'git apply' worked just fine, and never tried the cat/patch method: ================================= $ git show |head commit 30a2f3c60a84092c8084dfe788b710f8d0768cd4 Author: Linus Torvalds Date: Tue Aug 12 18:55:39 2008 -0700 Linux 2.6.27-rc3 diff --git a/Makefile b/Makefile index fd3ca6e..53bf6ec 100644 --- a/Makefile +++ b/Makefile $ git apply --check --verbose ../yh-patch1-2.6.27-rc3.diff Checking patch arch/x86/kernel/e820.c... Checking patch arch/x86/kernel/setup.c... Checking patch include/asm-x86/e820.h... $ git apply --verbose ../yh-patch1-2.6.27-rc3.diff Checking patch arch/x86/kernel/e820.c... Checking patch arch/x86/kernel/setup.c... Checking patch include/asm-x86/e820.h... Applied patch arch/x86/kernel/e820.c cleanly. Applied patch arch/x86/kernel/setup.c cleanly. Applied patch include/asm-x86/e820.h cleanly. dawitbro@fileserver:~/sandbox/git-kernel/linux-2.6$ git status # Not currently on any branch. # Changed but not updated: # (use "git add ..." to update what will be committed) # # modified: arch/x86/kernel/e820.c # modified: arch/x86/kernel/setup.c # modified: include/asm-x86/e820.h # no changes added to commit (use "git add" and/or "git commit -a") $ git diff > ../yh-patch1-2.6.27-rc3.git-apply.diff ================================= As you can see, I renamed your patch file to 'yh-patch1-2.6.27-rc3.diff', and after applying your patch I created a diff against 2.6.27-rc3 called 'yh-patch1-2.6.27-rc3.git-apply.diff'. I have attached the output of the following command so that you can see that your patch applied correctly: diff -y -W 200 yh-patch1-2.6.27-rc3.diff yh-patch1-2.6.27-rc3.git-apply.diff At this point, I retried a kernel with your patch... > can you try enable kexec and kdump in you .config. > > it should works. my .config have config_kexec I ran 'make oldconfig' to get my .config up to date, then 'make menuconfig' to make sure I had enabled CONFIG_KEXEC and CONFIG_CRASH_DUMP: $ egrep 'KEXEC|DUMP' .config CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y I took these steps, and posted this info, for the sake of our individual and collective sanity! ;-) I want you all to be sure that I applied the patch correctly, and adjusted the .config as requested, before building the kernel. The kernel built without error, so I installed and rebooted. It locked up in the usual way. I rebooted, using the "hpet=disable" parameter, and it booted just fine, just like all the others since 3def3d6d... >> I do not know how to bisect with your patch if I have a "bad" but no "good" >> to start with. Can you explain how I should proceed when I _do_ get home? >> (I can just enabled those config options and try the patch again, but I am >> confused about the bisect you are asking me to perform.) > > just like the old way doing git-bisect, but before compiling, apply > the batch, and before git-bisect good or bad, revert the patch. Since I am still confused about how I should perform the bisect you are asking for, I will wait until you can clarify. I responded earlier from work, where I had no access to a Linux machine, so I could not quote the git documentation I have read... to explain better why I am confused. Here are the sections of 'git help bisect' that I had in mind: ===== BEGIN QUOTED SECTIONS ================= [...] This command uses git-rev-list --bisect option to help drive the binary search process to find which change introduced a bug, given an old "good" commit object name and a later "bad" commit object name. [...] Basic bisect commands: start, bad, good The way you use it is: $ git bisect start $ git bisect bad # Current version is bad $ git bisect good v2.6.13-rc2 # v2.6.13-rc2 was the last version # tested that was good When you give at least one bad and one good versions, it will bisect the revision tree and say something like: Bisecting: 675 revisions left to test after this [...] and you continue along, compiling that one, testing it, and depending on whether it is good or bad, you say "git bisect good" or "git bisect bad", and ask for the next bisection. ===== END QUOTED SECTIONS ================= If there is a way to use 'git bisect' beginning with a "bad" version but no "good" version, then it is an advanced usage that I have not read about and do not understand how to use. As soon as you tell me how to carry out the process, I will do so and report the results. In the meantime, can you comment on the bisection I did last night? I found something very interesting about the commit that first causes the lockup (3def3d6d...), and the very next commit (1e934dda...) -- if I checkout 1e94... and try to revert the changes made in 3def..., the kernel freezes in spite of the revert. Because of this, I would conclude that your patch for 2.6.27-rc3 was doomed before you began, and we should look more carefully at the commits from February instead of trying to revert at the 2.6.27 HEAD. I am not a kernel developer, so my opinions are probably safe to ignore, but I think we should be trying to extract information from my (faulty) machines about what is happening differently between the "bad" commit 3def3d6d... and the "good" commit before it. I am too tired to continue my experiments tonight -- it is summer semester final exam time at the college where I tutor, and answer questions about calculus 2, linear algebra, nursing math, etc., has worn me right out -- but while I wait to find out the next step I should do with the 2.6.27-rc3 patch, I am going to try to get some useful info printed to the console before the kernel locks up back in those Feb. revisions. Thanks, Dave W.