Bug 777 - vesafb fails to initalize 1GB system memory and 128MB video memory
Summary: vesafb fails to initalize 1GB system memory and 128MB video memory
Status: CLOSED CODE_FIX
Alias: None
Product: Drivers
Classification: Unclassified
Component: Console/Framebuffers (show other bugs)
Hardware: i386 Linux
: P2 normal
Assignee: James Simmons
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2003-06-05 02:57 UTC by Christopher Rude
Modified: 2003-10-04 14:45 UTC (History)
1 user (show)

See Also:
Kernel Version: 2.5.63 - Current
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
the patch in patch form ;) (700 bytes, patch)
2003-06-05 02:59 UTC, Christopher Rude
Details | Diff

Description Christopher Rude 2003-06-05 02:57:06 UTC
Distribution: 
Gentoo Linux 1.4 profile w/ ARCH=~x86 
Hardware Environment: 
2xAthlon MP 1900+,Tyan 2462 (AMD 760MP), 1GB memory, VisionTek Gforce 4 Ti4600(128 
MB) 
Software Environment: 
GCC 3.2.3, Glibc 2.3.2 
Problem Description: 
trying to use vesafb, and get blank screen (system boots normaly w/ exception of no 
frambuffer) has something to do with failing to allopcate memory (ioremap i belive) this had 
been present in 2.4 series as well, untill recently (at least in the ac sources, and 
gentoo-sources) 
 
Steps to reproduce: install 1GB of memory, and use a video card w/ 128MB memory and try 
to use frambuffered console (I have seen behaviour on more than just my card. 
 
I have ported the patch that I had used in the 2.4 series (before gentoo and ac picked it up) 
 
please be kind as I am relativly new to this, and I hope this helps (btw I have been using this 
patch since 2.5.63, w/ slight change at 2.5.67, on linus's kernel and on the mm tree, and have 
had no troubles) 
 
______________________________________________ 
diff -Naur linux-2.5.67/drivers/video/vesafb.c 
linux-2.5.67-postpatch/drivers/video/vesafb.c 
--- linux-2.5.67/drivers/video/vesafb.c 2003-04-16 13:29:54.000000000 -0500 
+++ linux-2.5.67-postpatch/drivers/video/vesafb.c       2003-04-09 14:52:46.000000000 -0500 
@@ -225,7 +225,7 @@ 
        vesafb_defined.xres = screen_info.lfb_width; 
        vesafb_defined.yres = screen_info.lfb_height; 
        vesafb_fix.line_length = screen_info.lfb_linelength; 
-       vesafb_fix.smem_len = screen_info.lfb_size * 65536; 
+       vesafb_fix.smem_len = screen_info.lfb_width * screen_info.lfb_height * 
vesafb_defined.bits_per_pixel / 8; 
        vesafb_fix.visual   = (vesafb_defined.bits_per_pixel == 8) ? 
                FB_VISUAL_PSEUDOCOLOR : FB_VISUAL_TRUECOLOR; 
 
________________________________________________ 
 
please let me know what I can do better to make this easier on you.... 
 
Thanks for all your hard work, the 2.6 kernel is going to turn a lot of heads, been nothing but 
impressed with the developmental branch
Comment 1 Christopher Rude 2003-06-05 02:59:11 UTC
Created attachment 396 [details]
the patch in patch form ;)
Comment 2 Christopher Rude 2003-06-25 18:31:25 UTC
relevent dmesg output on vanilla 2.5.73 
 
grub kernel line: 
kernel /boot/bzImage.dev root=/dev/hda3 vga=794 video=vesa:1280x1024@70 
 
vesafb: abort, cannot ioremap video memory 0x8000000 @ 0xf0000000 
 
as you might supposes, still no framebuffer console present 
Comment 3 Christopher Rude 2003-06-25 19:55:53 UTC
relevent dmesg output on 2.5.73  w/ submitted patch 
 
vesafb: framebuffer at 0xf0000000, mapped to 0xf880a000, size 2560k 
vesafb: mode is 1280x1024x16, linelength=2560, pages=1 
vesafb: protected mode interface info at c000:e350 
vesafb: scrolling: redraw 
vesafb: directcolor: size=0:5:6:5, shift=0:11:5:0 
fb0: VESA VGA frame buffer device 
Console: switching to colour frame buffer device 160x64 
 
grub kernel line:  
kernel /boot/bzImage.dev root=/dev/hda3 vga=794 video=vesa:1280x1024@70 
 
and yes my framebuffer console works ;) 
Comment 4 Christopher Rude 2003-07-28 00:05:35 UTC
2.6 testing (i know test2, and test1-mm1) no longer suffer from the mentiond problem ;) 
 
good workd to whomever fixed the underlying issue 
 
closing is cool 
 
Thank you 
Comment 5 Lars G 2003-09-28 07:52:32 UTC
tried kernel 2.6 test4 and test5 and i still get the blank console problem.
tried the patch (had to disable the rivafb module, because of
some reference errors) and its working now.

this is with a geforce4 128mb and 1gig of ram.


lars
Comment 6 Lars G 2003-10-04 14:45:31 UTC
update

2.6-test6-bk4 works fine out of the box!!
thats really sweet!

lars


Note You need to log in before you can comment on or make changes to this bug.