Love-sources 2.6.0-test9 - Late but great

Message

Evangelion · Post by **Evangelion** » Mon Nov 17, 2003 1:10 pm

Lovechild wrote:
RedBeard0531 wrote:Lovechild: does the chang in avatar indicate your back on on gentoo? And how dare you wish death on the King of all DE's!
hat Ill stay here for a while.

2) because it sucks.

Offtopic: Why should KDE die? Seriously: why? Is someone forcing you to use it against your wishes? Is KDE somehow "ruining the Linux experience" for you? Like it or not, KDE has done KICK-ASS job on their desktop, and they have done ALOT to make Linux (and *BSD) more user-friendly! Just because you happen to use GNOME is not (IMO) a valid reason to hope that KDE would die. KDE provides the GNOME-folks a source of competition (and vice versa). The two teams learn from each other and try to outdo each other. That's one of the reason why the two projects have progressed so much in such a short time.

Honestly, I find the "I use XXXXX and I think every other piece of software besides the ones I use should just die!"-attitude to be extremely annoying.

I'm a KDE-user myself. Yet I wish no harm to Xfce, GNOME, Fluxbox, etc. etc. Far, far from it!

PrakashP · Post by **PrakashP** » Mon Nov 17, 2003 1:15 pm

[OT] Man, calm down. I am a KDE user myself, nevertheless I don't mind someone wishing KDE to die. Why must everybody be politically correct? Expressign oneself in an extreme way makes the world much more funny. My 2 cents[/OT]

Evangelion · Post by **Evangelion** » Mon Nov 17, 2003 1:33 pm

PrakashKC wrote:[OT] Man, calm down. I am a KDE user myself, nevertheless I don't mind someone wishing KDE to die. Why must everybody be politically correct? Expressign oneself in an extreme way makes the world much more funny. My 2 cents[/OT]

It would be 100% OK to say something like "I have tried both and I prefer GNOME". However, I find comments like "KDE in no shape or form causes any direct on indirect harm to me, but still, I wish that the project would just die! Everyone should use the software I use!" to be moronic in the extreme.

The reason for it to die? "because it sucks". How can it suck if it has thousands upon thousand of happy users and lots and lots of awards? KDE is one of the most popular and succesfull open source projects there are, how can it "suck" by any definition of the word?

antisthenes · Post by **antisthenes** » Mon Nov 17, 2003 1:38 pm

Well, Windows 9x sucks (I'm sure we would all agree), and that has millions of active users.

More OnT: I tried -love3....definitely not the same as -love2. I went back--I think I'll wait for -test10-love1.[/quote]

gen2newB · Post by **gen2newB** » Mon Nov 17, 2003 1:40 pm

i actually use both KDE and fluxbox. KDE when i want to type letters or administer accounts and fluxbox when i want a streamlined system without the frills loaded for maximum FPS in games =]. over all, i like KDE for its usability. Gnome seemed kind of kiddie to me, but that is just me.

GoinBald · Post by **GoinBald** » Mon Nov 17, 2003 1:51 pm

Thought I'd chime in here and Say :
THANK YOU!
Nary a hiccup with any load.
Great job!
U guys ROCK!

Humbly,
Baldy

bssteph · Post by **bssteph** » Mon Nov 17, 2003 2:13 pm

antisthenes wrote:More OnT: I tried -love3....definitely not the same as -love2. I went back--I think I'll wait for -test10-love1.

If you're up to it, please try unapplying the patches mentioned about 6 posts prior to this if you are using the ACPI power management timer. Doing so has helped other people.

(I'm going to ask for everyone to nix the DE/WM conversation for the sake of on-topicness... if you want to continue, make a "DE war: Love-sources spillover" in Gentoo Chat or something ;) It's valid to say "I don't like XYZ because of blah," but it's also OK [in my book at least] to say "XYZ sucks and should die" because both are subjective, just stated differently. And with that I'm done on this topic.)

PrakashP · Post by **PrakashP** » Mon Nov 17, 2003 3:33 pm

BTW, patched love3 using cfq really seems to fly!

i am very happy with this release now.

neenee · Post by **neenee** » Mon Nov 17, 2003 4:39 pm

patch love3 using cfq? is that the same as i am using
now? (love3 and elevator=cfq in grub.conf)

PrakashP · Post by **PrakashP** » Mon Nov 17, 2003 5:28 pm

No, I was referring to the apm-timer patches above, ie reverting them. Using this kernel with elevator=cfq really ROCKS.

neenee · Post by **neenee** » Mon Nov 17, 2003 5:53 pm

performance has decreased between love2 and love3
for me, though i use elevator=cfq. any ideas on how
to get that performance back are very welcome.

Lovechild · Post by **Lovechild** » Mon Nov 17, 2003 6:01 pm

neenee wrote:performance has decreased between love2 and love3
for me, though i use elevator=cfq. any ideas on how
to get that performance back are very welcome.

Read the thread

shm · Post by **shm** » Mon Nov 17, 2003 6:18 pm

a little OT, but has anyone had any difficulty in getting bootsplash configs working in test9 (specifically love2), that worked fine in 2.4? I can't get anything working besides the default.

THX!

die corn.

bssteph · Post by **bssteph** » Mon Nov 17, 2003 6:48 pm

shm wrote:a little OT, but has anyone had any difficulty in getting bootsplash configs working in test9 (specifically love2), that worked fine in 2.4? I can't get anything working besides the default.

Since it was fixed for grub I've only tried an initrd which I think I copied straight from /usr/share/bootsplash/
I haven't tried anything else.

shm wrote:die corn.

corn murderer!

Wedge_ · Post by **Wedge_** » Mon Nov 17, 2003 8:01 pm

Bootsplash just doesn't want to work for me. I've got all the right options selected in the kernel, as far as I can see, but if I try to set an image with "splash -s -u 0 ..." I get an error because /proc/splash ain't there. Is this normal on 2.6? I had bootsplash working on 2.4 months ago, but I haven't tried to use it since then. I don't get any bootsplash message during bootup either.

neenee · Post by **neenee** » Mon Nov 17, 2003 10:26 pm

Lovechild wrote:
neenee wrote:performance has decreased between love2 and love3
for me, though i use elevator=cfq. any ideas on how
to get that performance back are very welcome.
Read the thread

*bows head in shame*

sorry about that. i reversed those two
acpi patches, recompiled and rebooted.

i am not sure anything has changed.
i was not using apm nor acpi.

bssteph · Post by **bssteph** » Tue Nov 18, 2003 4:08 am

neenee wrote:
Lovechild wrote:
neenee wrote:performance has decreased between love2 and love3
for me, though i use elevator=cfq. any ideas on how
to get that performance back are very welcome.
Read the thread :)
*bows head in shame*

sorry about that. i reversed those two
acpi patches, recompiled and rebooted.

i am not sure anything has changed.
i was not using apm nor acpi.

LKML has shined a little bit more light on the situation...
http://www.ussg.iu.edu/hypermail/linux/ ... /0219.html
grep your dmesg for tsc, if you're not using it, try booting with the argument as seen at the url.
(X86_TSC is enabled automatically from the config, dependent on your processor)

Ricky · Post by **Ricky** » Tue Nov 18, 2003 7:10 am

Just tried out r3, no problems so far. And the slow mouse problem seems to have gone away.

(Not that it was a terribly major issue for me, a simple back-and-forth on my kvm switch made it go away, but that's one less thing I gotta do now

)

Corp.Nobbs · Post by **Corp.Nobbs** » Tue Nov 18, 2003 4:14 pm

@bssteph

Sorry, I'm gonna be a pain in the butt here..

I'm trying to track down where vmware falls over when you build under love3. Now I'm no programmer, so don't shoot me if I'm talking balls, but the error I get concerns a change in struct of do_mmap_pgoff.

I can grep change to this in love3:

root@athlon opt # cat 2.6.0-test9-love3.patch | grep do_mmap_pgoff
- error = do_mmap_pgoff(file, addr, len, prot, flags, pgoff);
+ error = do_mmap_pgoff(current->mm, file, addr, len, prot, flags, pgoff);
- error = do_mmap_pgoff(file, addr, len, prot, flags, pgoff);
+ error = do_mmap_pgoff(mm, file, addr, len, prot, flags, pgoff);
+ARC(mm_mmap_sem, mapping_i_shared_sem, "do_mmap_pgoff()");
-extern unsigned long do_mmap_pgoff(struct file *file, unsigned long addr,
+extern unsigned long do_mmap_pgoff(struct mm_struct *mm, struct file *file,
- ret = do_mmap_pgoff(file, addr, len, prot, flag, offset >> PAGE_SHIFT);
+ ret = do_mmap_pgoff(current->mm, file, addr, len, prot, flag,
-unsigned long do_mmap_pgoff(struct file * file, unsigned long addr,
+unsigned long do_mmap_pgoff(struct mm_struct *mm, struct file * file,

Which I'm supposing the change occurs. What I'm looking for, is which seperate patch contains these changes? I have downloaded all the patches listed in your post from page 3 of this thread and mm3 + updated nicks, and grep'ped each of these to findout which one I'm looking for, but I can't find a match

Could you shed some light

I'd like to then un-patch it and see if vmware compiles again.

PS: Those using vmware mods built from previous love's, word on the street is that as soon as vmware makes this do_mmap_pgoff call, the system will crash coz of the missmatch. I'm not sure when this would occur though..

bssteph · Post by **bssteph** » Tue Nov 18, 2003 4:29 pm

I would have to say that's reiser4 off the top of my head... I can't look at the broken out patches ATM... the reiser4 patch is a couple pages back if you want to unapply.

Corp.Nobbs · Post by **Corp.Nobbs** » Tue Nov 18, 2003 4:40 pm

bssteph wrote:I would have to say that's reiser4 off the top of my head... I can't look at the broken out patches ATM... the reiser4 patch is a couple pages back if you want to unapply.

Your da man!!

That's the cookie alright. I did grab the reiser4 stuff, but it was an earlier patch as linked in this thread, not the latest. So now, I'll try to remove it and see if vmware plays.. Thanks

Update: Yep it's the reiser4 patch pulling vmware down.. I un-applied the patch from love3 and now the vm modules compile OK. The only problem was that the reiser4 patch did not un-apply cleanly. I had to manually edit the following files as indicated by the reject files.
fs/sysfs/inode.c
include.linux/init_task.h
mm/Makefile

The reiser4 patch is here: http://thebsh.namesys.com/snapshots/200 ... r4.diff.gz

Thanks for the help bssteph.

bssteph · Post by **bssteph** » Tue Nov 18, 2003 9:22 pm

I forgot that there were failures in reiser4. This isn't a lot of help, but there's a bit about it in my notes.
http://bssteph.irtonline.org/linux/patc ... orklog.txt
None of the patch unapplies should have required anything fancy.

Dinini · Post by **Dinini** » Wed Nov 19, 2003 2:04 am

Wedge_ wrote:I get an error because /proc/splash ain't there. Is this normal on 2.6?

Odd, I get /proc/splash even if I don't use a initrd to set up the initial images.
You are using the vesafb in 16 bit mode right?
For the moment 16bit is the only supported color depth.

Just as a doublecheck, cd to your source directory, "make oldconfig" (just as a safeguard since it does some dependency checking) and "grep .config -e SPLASH -e PROC_FS". If you don't see "CONFIG_BOOTSPLASH=y" and "CONFIG_PROC_FS=y" then you have a kernel configuration problem.

antisthenes · Post by **antisthenes** » Wed Nov 19, 2003 7:53 am

-test9-mm4 is out. Merges the ACPI interactivity fix, some radeonfb updates, some general stuff. Looks like this is more substantive than -mm3.

Where is -test10, I wonder?

Anderas · Post by **Anderas** » Wed Nov 19, 2003 3:12 pm

Wedge_ wrote:Bootsplash just doesn't want to work for me. I've got all the right options selected in the kernel, as far as I can see, but if I try to set an image with "splash -s -u 0 ..." I get an error because /proc/splash ain't there. Is this normal on 2.6? I had bootsplash working on 2.4 months ago, but I haven't tried to use it since then. I don't get any bootsplash message during bootup either.

Wedge_, I am in the exact same situation as you. Bootsplash compiled in, but no /proc/splash, and no bootsplash. It worked back in 2.4. Someone please tell me how to fix this... :-(