Love-sources 2.6.0-test9 - Late but great

Message

bssteph · Post by **bssteph** » Fri Nov 07, 2003 2:25 am

And now with bootsplash (in love2)!

http://bssteph.irtonline.org/linux/patc ... .patch.bz2
http://bssteph.irtonline.org/linux/patc ... _notes.txt
http://bssteph.irtonline.org/linux/patc ... -r2.ebuild

Alrighty, a bunch of new stuff since I last checked in.
- bootsplash: needs some testing (so leave feedback)
- new versions abound
- some fixes
- reiser4 gone again

I haven't tested bootsplash (more on that later), but it patches cleanly and other people had luck with mm2, so give it a whirl.

Of course, check the notes, and That Disclaimer is still present.

Note: Lovechild released test9-love1 on the old thread here:
http://forums.gentoo.org/viewtopic.php?p=619481#619481
and the patch is mirrored here:
http://bssteph.irtonline.org/linux/patc ... -love1.bz2

UPDATES

test9-love3 is playing at a post near you.
test9-love4 found here

bssteph · Post by **bssteph** » Fri Nov 07, 2003 2:31 am

As I said, I'm not having -any- luck with bootsplash here. I've tried love2, mm2, and vanilla, all with the same problem and error.
So either I'm dumb or this machine doesn't like it... either way I'm going to try at home soon and we'll see where that goes. Again please post if someone has it working successfully in love2, I'd love to know. ;)

Addressing the multiple-threads idea, it's a good idea, and if enough want it (see the end of test8-love thread for details) we can do it, but for now you can just bookmark somewhere in my web directory to look for the latest patchset.
For example http://bssteph.irtonline.org/linux/patches/2.6/

IvoryTiger · Post by **IvoryTiger** » Fri Nov 07, 2003 2:53 am

Complied fine. Rebooting now.

Bootsplash didn't work...at first.

Played around with the command line and figured out why it didn't work for me.

I had vga=0x318 on my kernel line. Apparently bootsplash doesn't like that format. When I changed it to vga=791, I get the bootsplash in all it's glory with a bunch of extra bootsplash lines in the bootup and shutdown messages (trying to track that down now).

Everything looks great

bssteph · Post by **bssteph** » Fri Nov 07, 2003 6:02 am

Still no luck here. /sbin/splash sets the image just fine, and the image changes when init hits level 3, but nothing before.

But perhaps more pressing, from dmesg

Code: Select all

nvidia: no version magic, tainting kernel.
nvidia: module license 'NVIDIA' taints kernel.
0: nvidia: loading NVIDIA Linux x86 nvidia.o Kernel Module  1.0-4496  Wed Jul 16 19:03:09 PDT 2003
0: NVRM: AGPGART: unable to retrieve symbol table

Large draws (like during the start of kde) seem to be slow. DRM still works fine though.

paulisdead · Post by **paulisdead** » Fri Nov 07, 2003 7:30 am

Bootsplash is working great for me, and I'm using vga=0x31A for 1280x1024 Framebuffer with the vesa drivers. Gnome with Kahakai runs fine for me as well with the Nvidia drivers. I do get those extra bootsplash messages.

Zeddicus · Post by **Zeddicus** » Fri Nov 07, 2003 7:31 am

Hmm, kernel built and is running fine here, but boosplash doesn't want to do...anything. Getting late here, I'll play with it tomorrow.

My 1600x1200 framebuffer (vga=838) maybe at fault?

Maddog39 · Post by **Maddog39** » Fri Nov 07, 2003 8:50 am

I get these errors on compile(with bootsplash):

Code: Select all

drivers/built-in.o(.text+0xa75e3): In function `splash_getraw':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa7e67): In function `splash_verbose':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa837d): In function `splash_status':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa84aa): In function `splash_read_proc':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa868a): In function `splash_write_proc':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa8a4c): In function `splash_init':
: undefined reference to `initrd_start'
drivers/built-in.o(.text+0xa8a66): In function `splash_init':
: undefined reference to `initrd_end'

any ideas?

Corp.Nobbs · Post by **Corp.Nobbs** » Fri Nov 07, 2003 10:53 am

Maddog39 wrote:I get these errors on compile(with bootsplash):

Code: Select all

drivers/built-in.o(.text+0xa75e3): In function `splash_getraw':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa7e67): In function `splash_verbose':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa837d): In function `splash_status':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa84aa): In function `splash_read_proc':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa868a): In function `splash_write_proc':
: undefined reference to `con2fb_map'
drivers/built-in.o(.text+0xa8a4c): In function `splash_init':
: undefined reference to `initrd_start'
drivers/built-in.o(.text+0xa8a66): In function `splash_init':
: undefined reference to `initrd_end'

any ideas?

I got the same thing. I found that I had selected 'compiled-in fonts' as a module, so I switched this to yes and the compile completed.

I do have some sort of bootsplash, but it's kinda messy. I'll be looking through the bootsplash how-to forum link and see if I can get it to work.

neenee · Post by **neenee** » Fri Nov 07, 2003 11:17 am

emerging it now. compiling it soon after. i'll report back.

*update* it boots fine as is smooth as usual.

i did not try bootsplash (yet).

MrDooM · Post by **MrDooM** » Fri Nov 07, 2003 11:41 am

bootsplash 3.1.1-2003/11/02: looking for picture... no good signature found.

damn didn't work. I tried it with every possibility in menue.lst I know.
Any ideas?

OneOfOne · Post by **OneOfOne** » Fri Nov 07, 2003 12:01 pm

Illuminatus wrote:Hmm, kernel built and is running fine here, but boosplash doesn't want to do...anything. Getting late here, I'll play with it tomorrow.

My 1600x1200 framebuffer (vga=838) maybe at fault?

try 1280x1024 (0x31A)

peace

Corp.Nobbs · Post by **Corp.Nobbs** » Fri Nov 07, 2003 12:18 pm

MrDooM wrote:bootsplash 3.1.1-2003/11/02: looking for picture... no good signature found.

damn didn't work. I tried it with every possibility in menue.lst I know.
Any ideas?

Yep, me too. Haven't found a solution yet, but still looking...

MrDooM · Post by **MrDooM** » Fri Nov 07, 2003 12:55 pm

I tried mm2 Sources too patching for my own, but the same output. Didn't work for me. If anyone has an idea, pls post it.

Lovechild · Post by **Lovechild** » Fri Nov 07, 2003 1:08 pm

bssteph has everything down with the release process - except the sexy thread naming... we need funny thread naming.... it's half the fun of doing releases man...

GoinBald · Post by **GoinBald** » Fri Nov 07, 2003 1:16 pm

Hello Again Love-Crafters,

First AWWWWWWWW!
I wish I hadn't got to bed so early.

second, She's built but I gotta wait till i get home to boot it.

With regards to the "no good signature found." Have u updated bootsplash?
This was a prob with the utils earlier.(But then again it could be the caffeine talking)

FYI. I remember a post on mandrake bugs where This cropped up between mdk12-13 and disapeared after mdk18.
But I think it has to do with the fact that the image cannot be pulled from the initrd.

Has anyone had success with splash=silent?

One minor pita is the fact that I have to switch vts to get the thing to redraw.

Once again,
Thank you, *I gonna try to sneak out early today to boot this.*
Jesse

GoinBald · Post by **GoinBald** » Fri Nov 07, 2003 1:19 pm

Lovechild:
U got that right!
I also really like the grass roots feel here too.

Baldy

PrakashP · Post by **PrakashP** » Fri Nov 07, 2003 1:26 pm

I am getting this an make trying to put in bootsplash:

drivers/built-in.o(.text+0xa1059): In function `splash_init':
: undefined reference to `initrd_start'
drivers/built-in.o(.text+0xa1073): In function `splash_init':
: undefined reference to `initrd_end'

The above tip didn't work for me.

On what is bootsplash dependant?

BTW, I am currently debuggin as scheduler with Nick (if you follow lkml). SO, probably I can give you a fresh update, if he succedes. mm2's sheduler shows some abnomalities.

bssteph · Post by **bssteph** » Fri Nov 07, 2003 1:30 pm

Lovechild wrote:bssteph has everything down with the release process - except the sexy thread naming... we need funny thread naming.... it's half the fun of doing releases man...

lol, I'll give it a shot next time :D

For those with bootsplash issues, my core error is also the "no good signature found." deal.
I've been using the bootsplash in portage, 0.6-r2 (which as far as I can tell is the latest), if that's what you mean.

Wedge_ · Post by **Wedge_** » Fri Nov 07, 2003 1:38 pm

PrakashKC wrote:I am getting this an make trying to put in bootsplash:

drivers/built-in.o(.text+0xa1059): In function `splash_init':
: undefined reference to `initrd_start'
drivers/built-in.o(.text+0xa1073): In function `splash_init':
: undefined reference to `initrd_end'

The above tip didn't work for me. On what is bootsplash dependant?

It looks like you need initrd support (Device Drivers -> Block Devices -> [*] Initial RAM disk (initrd) support)

GoinBald · Post by **GoinBald** » Fri Nov 07, 2003 1:49 pm

bssteph:
Yup that's it.

From a posting I remember it had to do with concatenating
the image to the initrd.

Also what's your video= ,vga= string?

OT YA YA witty titles !

Thanks again u all ROCK,
Jesse

Lovechild · Post by **Lovechild** » Fri Nov 07, 2003 3:23 pm

You have to cat your image on the initrd uncompressed, or the bootsplash stuff with fail.

Zeddicus · Post by **Zeddicus** » Fri Nov 07, 2003 3:26 pm

Looking into it more closely, I also get the "no good signature error".

Using the vesa driver, vga=838 (tried with others, no effect).

Once the splash commands to the other ttys issue, I also get:

Code: Select all

bootsplash 3.1.1-2003/11/02: looking for picture.... found (1600x1200, 188314 bytes, v3)
bootsplash: status on console 1 changed to on

And so on, up to console 5. So, it seems to be doing fine in setting...just...doesn't DO anything

PrakashP · Post by **PrakashP** » Fri Nov 07, 2003 3:50 pm

@Wedge_

Thx, I'll try it.

@bssteph

So this seems to be the patch from Nick for the as scheduler to get it run nicely again.

Code: Select all


 linux-2.6-npiggin/drivers/block/as-iosched.c |    7 ++++---
 1 files changed, 4 insertions(+), 3 deletions(-)

diff -puN drivers/block/as-iosched.c~as-full drivers/block/as-iosched.c
--- linux-2.6/drivers/block/as-iosched.c~as-full	2003-11-08 01:40:49.000000000 +1100
+++ linux-2.6-npiggin/drivers/block/as-iosched.c	2003-11-08 02:04:45.000000000 +1100
@@ -962,7 +962,7 @@ static void as_completed_request(request
 	}
 
 	if (!blk_fs_request(rq))
-		return;
+		goto out;
 
 	if (ad->changed_batch && ad->nr_dispatched == 1) {
 		WARN_ON(ad->batch_data_dir == arq->is_sync);
@@ -982,7 +982,6 @@ static void as_completed_request(request
 	 * and writeback caches
 	 */
 	if (ad->new_batch && ad->batch_data_dir == arq->is_sync) {
-		WARN_ON(ad->batch_data_dir != REQ_SYNC);
 		update_write_batch(ad);
 		ad->current_batch_expires = jiffies +
 				ad->batch_expire[REQ_SYNC];
@@ -1508,8 +1507,10 @@ as_insert_request(request_queue_t *q, st
 
 	/* barriers must flush the reorder queue */
 	if (unlikely(rq->flags & (REQ_SOFTBARRIER | REQ_HARDBARRIER)
-			&& where == ELEVATOR_INSERT_SORT))
+			&& where == ELEVATOR_INSERT_SORT)) {
+		WARN_ON(1);
 		where = ELEVATOR_INSERT_BACK;
+	}
 
 	switch (where) {
 		case ELEVATOR_INSERT_BACK:

didl · Post by **didl** » Fri Nov 07, 2003 4:06 pm

Well,

Seems that my ATI mobility radeon 7500 does not like ati_dri-2.6.patch.
The radeon drivers fail to load and direct rendering is therefore
broken (worked like a charm before).
I get

Code: Select all

[drm] Initialized radeon 1.9.0 20020828 on minor 0
[drm:radeon_cp_init] *ERROR* radeon_cp_init called without lock held
[drm:radeon_unlock] *ERROR* Process 4262 using kernel context 0

Anybody have a solution of how to fix this?

bssteph: Would you be able to provide the broken out ati_dri-2.6.patch
so I can remove it for now.

Thanks, guys

bssteph · Post by **bssteph** » Fri Nov 07, 2003 5:19 pm

OOPS!
didl: I meant to fix that in love2 but forgot to throw the patch in my patching sandbox. I did the patch on my source so I totally missed the error (I have the same card).
Anyway the problem isn't in the new ati patch that I did add, it's a bug from mm2. Unapply this:
really long url - a patch from mm2