Daily blurbs Nov. 2004

Plans

daily blurbs


30 Nov 2004 (Tue)

08:09:22 # Life mmx-emu looks like an interesting library. It is probably LD_PRELOADed. It shows some techniques like handling SIGILL and decoding the current instruction to perform emulation. Would it be possible to do, for example, 486/386 emulation(theoretically)?

26 Nov 2004 (Fri)

08:08:16 # Life Update to soundtracker. I've noticed the topic on soundtracker on Mailinglist, and trying to update jack driver for soundtracker.

I've noticed that after applying the patch, things don't improve. I noticed that soundtracker is dying with a sigsegv somewhere when enabling jack output.

25 Nov 2004 (Thu)

08:55:06 # Life I had missing dependency in libecasound2.2-dev, which probably had gone unnoticed most of the year. Hmm... didn't I rebuild ecamegapedal often enough?

23 Nov 2004 (Tue)

09:32:36 # Life After investigating further on radeon performance issues in the new kernel, I am now suspecting fb driver change. It has somewhat changed.

11:07:42 # Life xine output log is not much different

--- log2.6.8	2004-11-21 00:03:49.000000000 +0900
+++ log2.6.9	2004-11-20 23:59:26.000000000 +0900
@@ -6,8 +6,8 @@
    ----------------------
         system name     : Linux
         node name       : ibookg4
-        release         : 2.6.8
-        version         : #1 Sat Sep 4 11:56:48 JST 2004
+        release         : 2.6.9
+        version         : #1 Fri Nov 12 09:13:41 JST 2004
         machine         : ppc
    CPU Informations:
    ----------------
@@ -201,42 +201,29 @@
 av_offset=0 pts
 spu_offset=0 pts
 ffmpeg_video_dec: direct rendering enabled
	

12:02:36 # Life Measured ssl speed on ibook.

OpenSSL 0.9.7e 25 Oct 2004
built on: Fri Nov 12 16:33:30 UTC 2004
options:bn(64,32) md2(char) rc4(idx,int) des(ptr,risc2,16,long) aes(partial) blowfish(idx)
compiler: gcc -fPIC -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -DOPENSSL_NO_KRB5 -DOPENSSL_NO_IDEA -DOPENSSL_NO_MDC2 -DOPENSSL_NO_RC5 -DB_ENDIAN -DTERMIO -O3 -fomit-frame-pointer -Wall
available timing options: TIMES TIMEB HZ=100 [sysconf value]
timing function used: times
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
md2                957.67k     2031.83k     2837.94k     3137.00k     3239.04k
mdc2                 0.00         0.00         0.00         0.00         0.00
md4               6547.32k    22540.30k    65129.84k   122686.74k   165716.97k
md5               5524.90k    18423.24k    50133.54k    87540.54k   112313.76k
hmac(md5)         7300.28k    23099.31k    58137.51k    93227.12k   113356.08k
sha1              5888.76k    19712.54k    52982.69k    92757.51k   117982.05k
rmd160            4748.82k    13838.06k    30774.26k    44149.85k    50957.70k
rc4              58834.21k    65191.05k    66637.97k    66751.87k    66981.82k
des cbc          20537.40k    22021.39k    22436.48k    22600.68k    22562.13k
des ede3          8057.81k     8279.63k     8364.27k     8379.08k     8350.19k
idea cbc             0.00         0.00         0.00         0.00         0.00
rc2 cbc          13942.81k    14603.24k    14794.06k    14865.66k    14850.12k
rc5-32/12 cbc        0.00         0.00         0.00         0.00         0.00
blowfish cbc     34270.32k    38769.85k    40034.34k    40327.59k    40380.09k
cast cbc         26653.17k    29408.62k    30052.02k    30298.39k    30287.80k
aes-128 cbc      33547.63k    35198.17k    35591.04k    35797.27k    35777.36k
aes-192 cbc      30443.81k    31871.89k    32284.87k    32330.95k    32383.82k
aes-256 cbc      27882.68k    29155.13k    29421.00k    29530.39k    29607.02k
                  sign    verify    sign/s verify/s
rsa  512 bits   0.0018s   0.0002s    571.2   5974.4
rsa 1024 bits   0.0092s   0.0005s    109.3   2023.7
rsa 2048 bits   0.0562s   0.0016s     17.8    613.1
rsa 4096 bits   0.3741s   0.0057s      2.7    174.7
                  sign    verify    sign/s verify/s
dsa  512 bits   0.0015s   0.0018s    662.8    555.5
dsa 1024 bits   0.0046s   0.0056s    216.2    179.0
dsa 2048 bits   0.0156s   0.0189s     64.2     52.9
	

20 Nov 2004 (Sat)

23:42:00 # Life After upgrading to 2.6.9, xine is much slower. I'm trying to figure out why it is the case. a diff to XFree86.0.log. There seems to be a DRM section that is being unloaded. PCI ids have apparently changed, but that doesn't affect the graphics card. dmesg for 2.6.8.

19 Nov 2004 (Fri)

22:09:35 # Life A ridiculous thread ridiculing my cat's page.

17 Nov 2004 (Wed)

08:14:44 # Life I've noticed that metronome stopped working in ecasound, and realized that it was due to not reading configuration file properly. Duh. That needs fixing.

16 Nov 2004 (Tue)

08:36:01 # Life I went through all bugs filed against pbuilder, and reviewed them. There are only about 30, that's not too bad.

14 Nov 2004 (Sun)

09:12:39 # Life I managed to remember to upload ecawave/ecamegapedal along with new version of ecasound, and while I was at it, updated to new release of ecamegapedal.

13 Nov 2004 (Sat)

11:39:39 # Life ecasound 2.3.5 was released, built a Debian package, and uploaded to experimental.

22:21:49 # Life TLUG meeting was held. Sacha gave a talk on PIM in general, and gave a plug about planner-el. It was nice to hear a talk on how to manage personal tasks, and why planner-el is suited. The best management is a software that does not require a visual/mental context switch to write down notes, and one which is aware of the context that one is trying to note down. Thus, planner-el knows what it's invoked from, and collects information from there; when reading email, when writing code, when browsing a webpage.

planner-el seems to have one drawback in that you need to be devoted to emacs, even for web browsing. I am using emacs for emails and code-writing, and diary-writing, but I browse the web with firefox, and IRC with xchat, and MSN through GAIM. That means, I cannot take the full advantage of planner-el. It would be ideal if I could have a shortcut key to note down my plan from firefox; and be able to add a note to my blog with the URL information and its title.

She demonstrated that planner-el is a massively integrated emacs masterpiece.

After the meeting, my hacking ideas are:

7 Nov 2004 (Sun)

09:17:03 # Life The relevant piece of code is this. Apparently it's tight-looping until READ_GET(pNv) is not same as Nv.dmaPut. It's probably waiting for DMA process to finish until proceeding to the next action. I am suspecting that it's waiting forever to reach a state that never happens.

void NVSync(ScrnInfoPtr pScrn)
{
    NVPtr pNv = NVPTR(pScrn);

    if(pNv->DMAKickoffCallback)
       (*pNv->DMAKickoffCallback)(pScrn);

    while(READ_GET(pNv) != pNv->dmaPut);

    while(pNv->PGRAPH[0x0700/4]);
}
	

16:54:02 # Life I've hacked up dancer-diary.el to add the day of the week. It's surprising that it's only a 15-minute hack and I never thought of doing it for the past 4 years.

6 Nov 2004

12:31:16 # Life X backtrace when X (this time, the debug server) stopped responding. This backtrace looks more sane to me.

This is probably related to this mail which points me to XFree86 Bugzilla Bug #1351.

[12:26:32]atoron:~> sudo gdb attach 1764
GNU gdb 6.1-debian
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i386-linux"...attach: No such file or directory.

Attaching to process 1764
Reading symbols from /usr/X11R6/bin/XFree86-debug...done.
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1".
Reading symbols from /usr/lib/libfreetype.so.6...done.
Loaded symbols for /usr/lib/libfreetype.so.6
Reading symbols from /usr/lib/libz.so.1...done.
Loaded symbols for /usr/lib/libz.so.1
Reading symbols from /lib/tls/i686/cmov/libm.so.6...done.
Loaded symbols for /lib/tls/i686/cmov/libm.so.6
Reading symbols from /lib/tls/i686/cmov/libc.so.6...done.
Loaded symbols for /lib/tls/i686/cmov/libc.so.6
Reading symbols from /lib/ld-linux.so.2...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /lib/tls/i686/cmov/libnss_files.so.2...done.
Loaded symbols for /lib/tls/i686/cmov/libnss_files.so.2
0x080debad in NVSync (pScrn=0x8b6b8b0) at nv_xaa.c:288
288	nv_xaa.c: No such file or directory.
	in nv_xaa.c
(gdb) bt
#0  0x080debad in NVSync (pScrn=0x8b6b8b0) at nv_xaa.c:288
#1  0x0837d5ab in XAACopyAreaFallback (pSrc=0x8ea2270, pDst=0x8dd2f80, 
    pGC=0x8d8ee28, srcx=0, srcy=0, width=476, height=34, dstx=386, dsty=0)
    at xaaFallback.c:81
#2  0x0837f0af in XAACopyArea (pSrcDrawable=0x8ea2270, pDstDrawable=0x8dd2f80, 
    pGC=0x8d8ee28, srcx=0, srcy=0, width=476, height=34, dstx=386, dsty=0)
    at xaaCpyArea.c:70
#3  0x0863b69f in miSpriteCopyArea (pSrc=0x8ea2270, pDst=0x8dd2f80, 
    pGC=0x8d8ee28, srcx=0, srcy=0, w=476, h=34, dstx=386, dsty=0)
    at misprite.c:1190
#4  0x084e13db in ProcCopyArea (client=0x8d68b38) at dispatch.c:1744
#5  0x084de350 in Dispatch () at dispatch.c:450
#6  0x084f58c4 in main (argc=9, argv=0xbffffda4, envp=0xbffffdcc) at main.c:469
(gdb) q
The program is running.  Quit anyway (and detach it)? (y or n) y
Detaching from program: /usr/X11R6/bin/XFree86-debug, process 1764
	

3 Nov 2004

11:02:04 # Life Debugging X server, I'm running 'nv' with noaccel. It seems to be very slow. My card is:

  Bus  1, device   0, function  0:
    VGA compatible controller: nVidia Corporation NV11DDR [GeForce2 MX 100 DDR/200 DDR] (rev 178).
      IRQ 11.
      Master Capable.  Latency=32.  Min Gnt=5.Max Lat=1.
      Non-prefetchable 32 bit memory at 0xde000000 [0xdeffffff].
      Prefetchable 32 bit memory at 0xd0000000 [0xd7ffffff].
	

I am suspecting that it's doing something wrong in its blitting function, but I'm not quite sure how I could catch the function just before the buffer overflow. I don't know a reliable way to reproduce the problem either.

11:29:44 # Life I'm trying -7 driver with acceleration now. I've noticed that Fabio has modified nv driver for -8, and -7. I should also try -6.

I've decided -7 is good. The diff between -7 and -8 are rather huge. It's almost impossible to figure out. A strategy for me is to work with -7 for the time being, and maybe investigate -8 problems when I have time, which is unlikely.

14:12:53 # Life I've finally bothered to upload new ecasound2.2 package to experimental. I'm planning to remove ecasound 2.0 series package and replace with ecasound2.2

2 Nov 2004

08:18:01 # Life X freezes on me since the recent dist-upgrade. Is 'nv' driver somewhat unstable? X just goes 100%.

277309 seems like a similar bugreport. I'm now checking if noaccel helps. NoAccel feels so much slower...

My wild guess is that there is a buffer overflow somewhere.

Script started on Tue Nov  2 08:51:10 2004
$ gdb 
GNU gdb 6.1-debian
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i386-linux".
(gdb) attach 30021
Attaching to process 30021
Reading symbols from /usr/X11R6/bin/XFree86...(no debugging symbols found)...done.
Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1".
Reading symbols from /usr/lib/libz.so.1...(no debugging symbols found)...done.
Loaded symbols for /usr/lib/libz.so.1
Reading symbols from /lib/tls/i686/cmov/libm.so.6...(no debugging symbols found)...done.
Loaded symbols for /lib/tls/i686/cmov/libm.so.6
Reading symbols from /lib/tls/i686/cmov/libdl.so.2...(no debugging symbols found)...done.
Loaded symbols for /lib/tls/i686/cmov/libdl.so.2
Reading symbols from /lib/libgcc_s.so.1...(no debugging symbols found)...done.
Loaded symbols for /lib/libgcc_s.so.1
Reading symbols from /lib/tls/i686/cmov/libc.so.6...(no debugging symbols found)...done.
Loaded symbols for /lib/tls/i686/cmov/libc.so.6
Reading symbols from /lib/ld-linux.so.2...(no debugging symbols found)...done.
Loaded symbols for /lib/ld-linux.so.2
Reading symbols from /lib/tls/i686/cmov/libnss_files.so.2...(no debugging symbols found)...done.
Loaded symbols for /lib/tls/i686/cmov/libnss_files.so.2
(no debugging symbols found)...0x083f7c63 in ?? ()
(gdb) bt
#0  0x083f7c63 in ?? ()
#1  0x086a0000 in ?? ()
#2  0x08682bd8 in ?? ()
#3  0xbffff7c8 in ?? ()
#4  0x08625b47 in ?? ()
#5  0x083f1528 in ?? ()
#6  0x00000000 in ?? ()
#7  0x00000001 in ?? ()
#8  0x00000001 in ?? ()
#9  0x00000001 in ?? ()
#10 0x00000000 in ?? ()
#11 0x00000054 in ?? ()
#12 0x00000000 in ?? ()
#13 0x00000000 in ?? ()
#14 0x00000000 in ?? ()
#15 0x081f10e0 in miBrokenRegion ()
#16 0x091501c8 in ?? ()
#17 0x09298028 in ?? ()
#18 0x09150368 in ?? ()
#19 0xbffff818 in ?? ()
#20 0x08627b3b in ?? ()
#21 0x09150368 in ?? ()
#22 0x09298028 in ?? ()
---Type <return> to continue, or q <return> to quit---
#23 0x091501c8 in ?? ()
#24 0x00000000 in ?? ()
#25 0x00000000 in ?? ()
#26 0x00000030 in ?? ()
#27 0x00000013 in ?? ()
#28 0x00000042 in ?? ()
#29 0x00000025 in ?? ()
#30 0x09150368 in ?? ()
#31 0x091437e0 in ?? ()
#32 0x0000007e in ?? ()
#33 0x00000000 in ?? ()
#34 0x00000006 in ?? ()
#35 0x00001882 in ?? ()
#36 0x0862c970 in ?? ()
#37 0x091501c8 in ?? ()
#38 0x09150270 in ?? ()
#39 0xbffff868 in ?? ()
#40 0x081947d9 in miSpriteInitialize ()
Previous frame inner to this frame (corrupt stack?)
(gdb) quit
The program is running.  Quit anyway (and detach it)? (y or n) y
Detaching from program: /usr/X11R6/bin/XFree86, process 30021
$ exit 

Script done on Tue Nov  2 08:51:33 2004
	

Junichi Uekawa

$Id: 200411.html.en,v 1.28 2004/11/29 23:10:59 dancer Exp $