<?xml version="1.0" encoding="UTF-8" standalone="yes" ?>
<!DOCTYPE bugzilla SYSTEM "https://bugzilla.yoctoproject.org/page.cgi?id=bugzilla.dtd">

<bugzilla version="5.0.6"
          urlbase="https://bugzilla.yoctoproject.org/"
          
          maintainer="it-coreprojects-helpdesk@linuxfoundation.org"
>

    <bug>
          <bug_id>363</bug_id>
          
          <creation_ts>2010-09-29 19:36:30 +0000</creation_ts>
          <short_desc>dmesg error from uvesafb with qemux86 with sato image</short_desc>
          <delta_ts>2010-12-21 06:07:17 +0000</delta_ts>
          <reporter_accessible>1</reporter_accessible>
          <cclist_accessible>1</cclist_accessible>
          <classification_id>7</classification_id>
          <classification>Build System, Metadata &amp; Runtime</classification>
          <product>Meta-yocto</product>
          <component>meta-yocto</component>
          <version>0.9</version>
          <rep_platform>x86</rep_platform>
          <op_sys>x86</op_sys>
          <bug_status>VERIFIED</bug_status>
          <resolution>FIXED</resolution>
          
          
          <bug_file_loc></bug_file_loc>
          <status_whiteboard></status_whiteboard>
          <keywords></keywords>
          <priority>Medium</priority>
          <bug_severity>normal</bug_severity>
          <target_milestone>0.9 M4</target_milestone>
          
          
          <everconfirmed>1</everconfirmed>
          <reporter name="xianchao zhang">xianchao.zhang</reporter>
          <assigned_to name="Edwin Zhai">edwin.zhai</assigned_to>
          <cc>edwin.zhai</cc>
    
    <cc>jessica.zhang</cc>
    
    <cc>jiajun.xu</cc>
    
    <cc>ke.yu</cc>
    
    <cc>kevin.tian</cc>
    
    <cc>poky.bs.watcher</cc>
    
    <cc>poky.watcher</cc>
    
    <cc>richard.purdie</cc>
    
    <cc>sgw</cc>
    
    <cc>xianchao.zhang</cc>
          
          
          <cf_os>---</cf_os>
          <cf_regression_type>---</cf_regression_type>
          <cf_verified>Program Management</cf_verified>
          <cf_docchange>---</cf_docchange>

      

      

      

          <comment_sort_order>oldest_to_newest</comment_sort_order>  
          <long_desc isprivate="0" >
    <commentid>10054</commentid>
    <comment_count>0</comment_count>
    <who name="xianchao zhang">xianchao.zhang</who>
    <bug_when>2010-09-29 19:36:30 +0000</bug_when>
    <thetext>Testing Env:    
Host: Ubuntu 10.04 LTS
Poky gitinfo(used to build qemu-native):
8976fa69d0ddfdad24fe041f4b88c853dc7a33a2
Qemuarm images: sato from nightly build on autobuilder,
http://autobuilder.pokylinux.org/nightly/20100928-1/

Steps:

1. download the sato image from autobuilder
2. boot up qemux86 with sato image
3. run command &apos;dmesg | grep -i error&apos;, the error log arise.

uvesafb: probe of uvesafb.0 failed with error -22</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10110</commentid>
    <comment_count>1</comment_count>
    <who name="Darren Hart">dvhart</who>
    <bug_when>2010-09-30 15:34:16 +0000</bug_when>
    <thetext>Guessing this is related to Bug 330. Will confirm next week.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10184</commentid>
    <comment_count>2</comment_count>
    <who name="Darren Hart">dvhart</who>
    <bug_when>2010-10-06 15:53:46 +0000</bug_when>
    <thetext># dmesg | grep vesa
uvesafb: failed to execute /sbin/v86d
uvesafb: make sure that the v86d helper is installed and executable
uvesafb: Getting VBE info block failed (eax=0x4f00, err=-2)
uvesafb: vbe_init() failed with -22
uvesafb: probe of uvesafb.0 failed with error -22

#cat /proc/cmdline 
vga=0 root=/dev/hda mem=128M ip=192.168.7.2::192.168.7.1:255.255.255.0

Notice that there is not a valid video= line, this may be part of the problem. As for the v86d helper application, I suspect it is not being added to the initrd filesystem.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10185</commentid>
    <comment_count>3</comment_count>
    <who name="Darren Hart">dvhart</who>
    <bug_when>2010-10-06 16:17:20 +0000</bug_when>
    <thetext>Hrm, I&apos;m not seeing an initramfs file in the booted fs, and one isn&apos;t specified on the qemu command line. Is an initramfs used?

Thinking the *.rootfs.* files might form the initramfs, I checked both the tar.gz2 and the .ext3 image, both contain /sbin/v86d:

$ sudo mount poky-image-sato-qemux86-20101006213825.rootfs.ext3 mnt -o loop
$ file mnt/sbin/v86d mnt/sbin/v86d: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.4.0, stripped

Can someone explain to me which files are used for what with a qemu86 target, and whether or not we use an initramfs?</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10226</commentid>
    <comment_count>4</comment_count>
    <who name="Kevin Tian">kevin.tian</who>
    <bug_when>2010-10-07 18:38:03 +0000</bug_when>
    <thetext>(In reply to comment #3)
&gt; Hrm, I&apos;m not seeing an initramfs file in the booted fs, and one isn&apos;t specified
&gt; on the qemu command line. Is an initramfs used?
&gt; Thinking the *.rootfs.* files might form the initramfs, I checked both the
&gt; tar.gz2 and the .ext3 image, both contain /sbin/v86d:
&gt; $ sudo mount poky-image-sato-qemux86-20101006213825.rootfs.ext3 mnt -o loop
&gt; $ file mnt/sbin/v86d mnt/sbin/v86d: ELF 32-bit LSB executable, Intel 80386,
&gt; version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.4.0,
&gt; stripped
&gt; Can someone explain to me which files are used for what with a qemu86 target,
&gt; and whether or not we use an initramfs?

It looks that qemux86 doesn&apos;t use a initramfs. v86d is designed to work as a module. fbsetup is the script to modprobe uvesafb.ko which is expected to run just before psplash. Since wr kernel has uvesafb built in the kernel, it fails earlier due to no initramfs, and also latter since no module is available. It may be the reason, but still need verification</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10232</commentid>
    <comment_count>5</comment_count>
    <who name="Kevin Tian">kevin.tian</who>
    <bug_when>2010-10-07 23:07:40 +0000</bug_when>
    <thetext>well, after changing uvesafb to a module, the original issue is gone. But there&apos;s still mmap error on framebuffer. Edwin is looking at it now.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10242</commentid>
    <comment_count>6</comment_count>
    <who name="Kevin Tian">kevin.tian</who>
    <bug_when>2010-10-08 01:43:45 +0000</bug_when>
    <thetext>the new error &quot;cannot mmap framebuffer&quot; error comes from psplash, which gets a zero stride. Further trace shows that bogus stride value comes from uvesafb module, which may further come from v86d emulation when uvesafb tries to get the mode info. Not sure whether it&apos;s related to my debug environment, where I manually rebuild linux-libc-header and v86d to catch up connector ring size fix by Edwin...</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10278</commentid>
    <comment_count>7</comment_count>
    <who name="xianchao zhang">xianchao.zhang</who>
    <bug_when>2010-10-09 01:12:33 +0000</bug_when>
    <thetext>With the new build images (20101007), there arise new segfault.

Steps:
1. download the build image
2. boot up target with sato/sdk image
3. run command &apos;dmesg | grep -i error&apos;

qemux86+sato:
vesafb: probe of uvesafb.0 failed with error -22
tasks[1690]: segfault at 0 ip 44255311 sp bfff45dc error 4 in libc-2.12.so[441e5000+13d000]

qemux86+sdk:
uvesafb: probe of uvesafb.0 failed with error -22
e-addressbook-f[1994]: segfault at 170 ip b763fb8c sp b7623090 error 4 in libebookbackendfile.so[b763d000+7000]
contacts[1991]: segfault at 4 ip 45532a49 sp bfe1fadc error 4 in libglib-2.0.so.0.2400.1[45520000+c1000]</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10279</commentid>
    <comment_count>8</comment_count>
    <who name="Edwin Zhai">edwin.zhai</who>
    <bug_when>2010-10-09 01:21:47 +0000</bug_when>
    <thetext>Xianchao,

The uvesafb issue will be fixed soon.
New segfault comes from contacts and tasks. Can you check if they can start? If not pls. open new bugs for them.

Thanks,
edwin</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10281</commentid>
    <comment_count>9</comment_count>
    <who name="xianchao zhang">xianchao.zhang</who>
    <bug_when>2010-10-09 01:44:17 +0000</bug_when>
    <thetext>(In reply to comment #8)
&gt; Xianchao,
&gt; 
&gt; The uvesafb issue will be fixed soon.
&gt; New segfault comes from contacts and tasks. Can you check if they can start? If
&gt; not pls. open new bugs for them.
&gt; 
&gt; Thanks,
&gt; edwin

contacts can start up, however tasks can not, I have opened a bug for this, bug 420.

Thanks!
xianchao</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10282</commentid>
    <comment_count>10</comment_count>
    <who name="Edwin Zhai">edwin.zhai</who>
    <bug_when>2010-10-09 01:57:29 +0000</bug_when>
    <thetext>Many thanks for Kevin to root cause and fix it.
mmap in psplash failed due to the length is 0, which caused by fb-&gt;stride is 0.

1: fb-&gt;stride comes from fb_info:fix.line_length via ioctl(FBIOGET_VSCREENINFO). 

2: line_length is set as mode-&gt;bytes_per_scan_line in drivers/video/uvesafb.c::uvesafb_set_par.

3: all available modes are get in uvesafb_vbe_getmodes with the user space helper v86d.

Although 1&amp; 3 works well, 2 is not called as the calling is from fbcon.c and FRAMEBUFFER_CONSOLE is not configured in wrs kernel.

Previous kernel configure it, but wrs kernel disable it so we have psplash failure for a long time.

make CONFIG_FRAMEBUFFER_CONSOLE=y in wrs kernel can fix it.</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10322</commentid>
    <comment_count>11</comment_count>
    <who name="Saul Wold">sgw</who>
    <bug_when>2010-10-11 11:19:14 +0000</bug_when>
    <thetext>http://git.pokylinux.org/cgit/cgit.cgi/poky/commit/?id=efe55d749ac0201d32587d94670fea46c8a98728</thetext>
  </long_desc><long_desc isprivate="0" >
    <commentid>10365</commentid>
    <comment_count>12</comment_count>
    <who name="Jiajun Xu">jiajun.xu</who>
    <bug_when>2010-10-11 22:56:46 +0000</bug_when>
    <thetext>Verified with M4 20101011 build, the bug is fixed.</thetext>
  </long_desc>
      
      

    </bug>

</bugzilla>