Menu
▾
▴
dprobes-list
You can subscribe to this list here.
| 2000 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
(1) |
Sep
(1) |
Oct
(5) |
Nov
(18) |
Dec
(10) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2001 |
Jan
(4) |
Feb
|
Mar
(11) |
Apr
(11) |
May
(2) |
Jun
(7) |
Jul
(14) |
Aug
(2) |
Sep
|
Oct
(10) |
Nov
(31) |
Dec
(13) |
| 2002 |
Jan
(21) |
Feb
(22) |
Mar
(31) |
Apr
(20) |
May
(5) |
Jun
(18) |
Jul
(3) |
Aug
(17) |
Sep
(16) |
Oct
(47) |
Nov
(35) |
Dec
(16) |
| 2003 |
Jan
(15) |
Feb
(8) |
Mar
(14) |
Apr
(26) |
May
(31) |
Jun
(15) |
Jul
(13) |
Aug
(4) |
Sep
|
Oct
(1) |
Nov
(2) |
Dec
(10) |
| 2004 |
Jan
|
Feb
(3) |
Mar
(4) |
Apr
(2) |
May
(2) |
Jun
(3) |
Jul
(2) |
Aug
(28) |
Sep
(7) |
Oct
(24) |
Nov
(27) |
Dec
(15) |
| 2005 |
Jan
(4) |
Feb
(4) |
Mar
(1) |
Apr
(8) |
May
(1) |
Jun
(1) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(2) |
Dec
|
| 2006 |
Jan
|
Feb
(3) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
| 2007 |
Jan
|
Feb
|
Mar
|
Apr
(4) |
May
(1) |
Jun
|
Jul
(1) |
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Cheri S. <Cr...@gm...> - 2007-05-15 04:47:16
|
<html><head><title>twinge</title></head> <body> Ferdinand,<br><br> <a href="http://0x000000D8.0000000323.000000000236.0x0000000000007D">267322 20</a><br><br> Corinne </body></html> |
|
From: Ruta G. <Rut...@am...> - 2007-04-29 21:08:43
|
ANLEGER UHR BJ5N.F!!! DIE RALLYE IST GESTARTET Firma: BOERSE INVEST BETEI WKN : 797639 ISIN : CH0012802093 Markt: Frankfurt Kürzel : BJ5N.F Preis: 1.90 5-Tag Prognose: 3.00 KAUFEN KAUFEN KAUFEN! BJ5N.F ESGESCHAFT FIN UNTER PARI! |
|
From: Linda C. <Aar...@gm...> - 2007-04-29 05:38:52
|
<html><head><title>singes</title></head> <body> Ellis,<br><br> <a href="http://0x000000000D8.00000000000323.000000000000000236.0x000000000 000007C">50411427</a><br><br> Charles </body></html> |
|
From: Liam M. <Pan...@gm...> - 2007-04-28 12:10:06
|
<html><head><title>softwares</title></head> <body> Dona,<br><br> <a href="http://0x00000000000D8.00000000000323.0x000000000000000000000009E. 124">40210768</a><br><br> Drake </body></html> |
|
From: maja H. <maj...@GE...> - 2007-04-25 18:29:49
|
without |
|
From: Roland C. <Liq...@gm...> - 2006-10-24 01:42:49
|
reparation directrix Up On the News - Q4 Demand Predicted to Remain Strong for Platinum IT Consulting Target Niche compton. Oracle, SQL Server and Sybase Can't All Be Wrong. Staffing companies are the #1 employer in the USA today, and IT staffing with GITH is leading the revolution. essay. NEW YORK, NY (PRIMEZONE) -- GITH -- A recent report based on a survey of fourteen hundred Chief Information Officers predicts steady growth in Q4 for technology IT hiring. Thirty-four percent of CIOs surveyed plan to add IT staff and none anticipated cutbacks in personnel. integer. GITH sources personnel for companies and offers custom software solutions related to Oracle, SQL Server and Sybase and software such as SAP, JD Edwards, and PeopleSoft. The report done on behalf of Robert Half Technology specifically mentions that 71% of CIOs anticipated needing personnel having skill sets related to Oracle, SQL Server, and DB2. quillwort. Read the entire news release misanthropic. http://biz.yahoo.com/pz/061020/107201.html GITH has been releasing steady news worldwide, from Yahoo Finance & Bloomberg to MSN & CNN Money, even the NASDAQ. Exposure for GITH is expansive. The increased frequency of news leads us to believe that something big is coming for GITH. mesozoic. Oct. 17, 2006 (PRIMEZONE) -- Executive Job Market Report Forecasting 'War For Talent' Bolsters Comments by GITH President lathe. http://news.moneycentral.msn.com/ticker/article.asp?Symbol=US:GITH&Feed=PZ&Date=20061017&ID=6111313 Oct. 14, 2006 (PRIMEZONE) -- GITH Foresees Growth in Fourth Quarter... posts increase in sales revenue... kitchenette. http://www.bloomberg.com/apps/news?pid=conewsstory&refer=conews&tkr=GITH:US&sid=aMMgvBh5UWMA Sept. 25, 2006 (PRIMEZONE) -- Market forces combining over the next few years to drive up demand for a decreasing supply of specialized IT personnel. GITH is leading the 21st Century revolution... achieve. http://bigcharts.marketwatch.com/news/articles.asp?guid={F1A16E67-50AD-4860-85E5-8FCC85B1B5B9}&newsid=882141976&symb=GITH&sid=2384270 horsetail concourse monroe |
|
From: Richard J M. <ric...@uk...> - 2006-02-21 08:35:06
|
The dprobes record header includes:
major, minor, cpuid (if SMP), pid, uid, cs:eip, ss:esp, timestamp, proc=
name
see dprobes/i386/dprobes.c and include/asm/dprobes.h
- -
Richard J Moore
IBM Advanced Linux Response Team - Linux Technology Centre
MOBEX: 264807; Mobile (+44) (0)7739-875237
Office: (+44) (0)1962-817072
=
Vara Prasad =
<pra...@us... =
m.com> =
To
Sent by: Andrey <di...@gm...> =
dprobes-list-ad =
cc
mi...@li... dpr...@li....n=
et
eforge.net =
bcc
=
Subj=
ect
20/02/2006 Re: [Dprobes-list] DProbes logs =
to
22:40 LTT 56 bytes =
=
=
Hi Andrey,
I am glad to know that you are trying to use dprobes to solve your
problem but most of the development has now moved from dprobes to
systemtap. Systemtap is a feature which is much richer and easier to us=
e
than Dprobes.
You can find out all about systemtap at http://www.sourceware.org/syste=
mtap
Please let me know if you have any questions about systemtap.
Thanks,
Vara Prasad
Andrey wrote:
> Hello...
> I use LTT 0.9.5a and DProbes 3.6.4 on linux kernel 2.4.19
>
> And I can see the following event in LTT log file:
>
>/usr/src/linux/vmli 1,134,583,116,745,927 N/A 75
>01 00 38 00 FB 07 00 00 07 00 6E 02 CA 08 00 00 00 00 00 00 10 00 F0
>8C 25 C0 10 00 84 92 10 C0 4C 5D A0 43 C6 61 0B 00 70 69 6E 67 00 6C
>6C 00 00 00 00 00 00 00 00 00
>
> Especially, I'm very interested in Description field. Experimenting,
>I found that 12 & 11 bytes are the "minor" value. The last 16 bytes
>seem to be the process name. So, 026Eh =3D 622d is a minor value, and
>"70 69 6E 67 00" =3D "ping" is a process name.
> But what do the other bytes mean???
>
>P.S. I use the following commands:
># dprobes -i myprobes.rpn -s /usr/src/linux/System.map -l ALL --log-ta=
rget
LTT
># tracedaemon -eCSTM -eNEWEV -b10000000 -ts15 /dev/tracer out.trace
out.proc
>
>Waiting for anybody's reply...
>_______
> Andrey
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: Splunk Inc. Do you grep through log=
files
>for problems? Stop! Download the new AJAX search engine that makes
>searching your log files as easy as surfing the web. DOWNLOAD SPLUNK=
!
>http://sel.as-us.falkag.net/sel?cmd=3Dk&kid=103432&bid#0486&dat=121642=
>_______________________________________________
>Dprobes-list mailing list
>Dpr...@li...
>https://lists.sourceforge.net/lists/listinfo/dprobes-list
>
>
>
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log
files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!=
http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=
=3D121642
_______________________________________________
Dprobes-list mailing list
Dpr...@li...
https://lists.sourceforge.net/lists/listinfo/dprobes-list
=
|
|
From: Vara P. <pra...@us...> - 2006-02-20 22:40:59
|
Hi Andrey, I am glad to know that you are trying to use dprobes to solve your problem but most of the development has now moved from dprobes to systemtap. Systemtap is a feature which is much richer and easier to use than Dprobes. You can find out all about systemtap at http://www.sourceware.org/systemtap Please let me know if you have any questions about systemtap. Thanks, Vara Prasad Andrey wrote: > Hello... > I use LTT 0.9.5a and DProbes 3.6.4 on linux kernel 2.4.19 > > And I can see the following event in LTT log file: > >/usr/src/linux/vmli 1,134,583,116,745,927 N/A 75 >01 00 38 00 FB 07 00 00 07 00 6E 02 CA 08 00 00 00 00 00 00 10 00 F0 >8C 25 C0 10 00 84 92 10 C0 4C 5D A0 43 C6 61 0B 00 70 69 6E 67 00 6C >6C 00 00 00 00 00 00 00 00 00 > > Especially, I'm very interested in Description field. Experimenting, >I found that 12 & 11 bytes are the "minor" value. The last 16 bytes >seem to be the process name. So, 026Eh = 622d is a minor value, and >"70 69 6E 67 00" = "ping" is a process name. > But what do the other bytes mean??? > >P.S. I use the following commands: ># dprobes -i myprobes.rpn -s /usr/src/linux/System.map -l ALL --log-target LTT ># tracedaemon -eCSTM -eNEWEV -b10000000 -ts15 /dev/tracer out.trace out.proc > >Waiting for anybody's reply... >_______ > Andrey > > >------------------------------------------------------- >This SF.net email is sponsored by: Splunk Inc. Do you grep through log files >for problems? Stop! Download the new AJAX search engine that makes >searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! >http://sel.as-us.falkag.net/sel?cmd=k&kid3432&bid#0486&dat1642 >_______________________________________________ >Dprobes-list mailing list >Dpr...@li... >https://lists.sourceforge.net/lists/listinfo/dprobes-list > > > |
|
From: Andrey <di...@gm...> - 2006-02-15 22:27:21
|
Hello... I use LTT 0.9.5a and DProbes 3.6.4 on linux kernel 2.4.19 And I can see the following event in LTT log file: /usr/src/linux/vmli 1,134,583,116,745,927 N/A 75 01 00 38 00 FB 07 00 00 07 00 6E 02 CA 08 00 00 00 00 00 00 10 00 F0 8C 25 C0 10 00 84 92 10 C0 4C 5D A0 43 C6 61 0B 00 70 69 6E 67 00 6C 6C 00 00 00 00 00 00 00 00 00 Especially, I'm very interested in Description field. Experimenting, I found that 12 & 11 bytes are the "minor" value. The last 16 bytes seem to be the process name. So, 026Eh =3D 622d is a minor value, and "70 69 6E 67 00" =3D "ping" is a process name. But what do the other bytes mean??? P.S. I use the following commands: # dprobes -i myprobes.rpn -s /usr/src/linux/System.map -l ALL --log-target = LTT # tracedaemon -eCSTM -eNEWEV -b10000000 -ts15 /dev/tracer out.trace out.pro= c Waiting for anybody's reply... _______ Andrey |
|
From: Frank C. E. <fc...@re...> - 2005-11-30 19:51:26
|
Hi - On Wed, Nov 30, 2005 at 02:31:03PM -0500, Devon Miller wrote: > [...] > Is Dprobes dead? Or, has it been replaced by something else? I believe most effort is now focused on systemtap: <http://sourceware.org/systemtap/>. - FChE |
|
From: Devon M. <dev...@gm...> - 2005-11-30 19:31:12
|
The last posting was in June and the last release was way before that. Is Dprobes dead? Or, has it been replaced by something else? |
|
From: Gettert W. (e. C. MD C. TS S. <Wol...@mc...> - 2005-06-21 18:14:31
|
Hi all, because of sopme priority shifting I am currently not working on the ARM port anymore. Since there seem to be some interest in an ARM port of kprobes I will post my first results here. What's working: - register/unregister a kprobe on an Sgold procressor - fire pre-, fault- and posthandler. What's not working: - Cleanup of registers and stack if the probed instruction is a branch or so ... - Thumb mode is not supported - Java mode is not supported The patch is against a kernel 2.6.9 however not a vanilly one since there is no sgold support in the mainstream kernel. But I think if there are dependencies to Sgold they can be easily fixed. I looked ober the patch and I didn't see something in the first glance. If you have any questions feel free to ask. Wolfram |
|
From: Andrey <di...@gm...> - 2005-05-12 01:04:31
|
Hello, all I've successfully installed LTT and DProbes and now they both can work at the same time with the same kernel. So, here are commands I usually type: insmod tracer insmod dp dprobes -i probes.rpn -s /usr/src/linux/System.map -l ALL --log-target LTT tracedaemon -eCSTM -eNEWEV -b10000000 -ts15 /dev/tracer out.trace out.proc traceview out But the problem is that TraceVisualizer doesn't build a graph :( (Says smth like not enough information). If I delete "-eCSTM -eNEWEV" then it's all ok, but I'm only interested in DProbes events, so I need these options. Can anyone tell me how make LTT build a graph with the information taken from DProbes? DProbes v3.6.4 LTT 0.9.5a kernel 2.4.19 (RedHat 8.0) Sample probes.rpn: // *** File Header *** name="/usr/src/linux/vmlinux" modtype=kernel major=7 id=0 logmax=1024 jmpmax=256 logonfault=no // *** Probe Point Definitions *** address=tcp_destroy_sock opcode=0x55 minor=1 exit address=fn_hash_lookup opcode=0x55 minor=2 exit address=inet_create opcode=0x55 minor=3 exit address=inet_release opcode=0x55 minor=4 exit address=sock_close opcode=0x55 minor=8 exit address=sock_create opcode=0x55 minor=9 exit address=ip_rcv opcode=0x55 minor=32 exit address=ip_forward opcode=0x55 minor=44 exit address=ip_route_input opcode=0x55 minor=45 exit ... _______ Andrey |
|
From: Tom Z. <za...@us...> - 2005-04-28 13:51:36
|
Peter Lauterbach writes: > I am attempting to compile a kprobes example from Tom Z. : > "syscall-kprobes.tar.bz2" on my SLES9-SP1 box. You probably need to either apply the kprobes/jprobes patch, or use a more recent kernel (I think kprobes was included in mainline around 2.6.10 or so). There's a pointer to the kprobes patches I used for this example in my original posting: http://www.listserv.shafik.org/pipermail/ltt-dev/2004-August/000649.html Tom |
|
From: Peter L. <p.l...@co...> - 2005-04-28 03:46:18
|
I am attempting to compile a kprobes example from Tom Z. :
"syscall-kprobes.tar.bz2" on my SLES9-SP1 box.
linux:~/work/syscall-kprobes # make
make -C /lib/modules/2.6.5-7.139-default/build SUBDIRS=`pwd` modules
make[1]: Entering directory `/usr/src/linux-2.6.5-7.139'
CC [M] /root/work/syscall-kprobes/trace-unlink-params.o
/root/work/syscall-kprobes/trace-unlink-params.c:28: error: field `jp'
has incomplete type
/root/work/syscall-kprobes/trace-unlink-params.c: In function `jsys_unlink':
/root/work/syscall-kprobes/trace-unlink-params.c:39: warning: implicit
declaration of function `jprobe_return'
/root/work/syscall-kprobes/trace-unlink-params.c: In function
`init_scp_trace':
/root/work/syscall-kprobes/trace-unlink-params.c:51: warning: implicit
declaration of function `register_jprobe'
/root/work/syscall-kprobes/trace-unlink-params.c: In function
`cleanup_scp_trace':
/root/work/syscall-kprobes/trace-unlink-params.c:63: warning: implicit
declaration of function `unregister_jprobe'
make[2]: *** [/root/work/syscall-kprobes/trace-unlink-params.o] Error 1
make[1]: *** [_module_/root/work/syscall-kprobes] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.5-7.139'
make: *** [build] Error 2
in the example, it is complaining about:
#include <linux/module.h>
#include <linux/kprobes.h>
#include <linux/kallsyms.h>
#include <linux/ltt.h>
static struct scp_trace_obj {
const char *funcname;
struct jprobe jp;
int trace_id;
} scpt;
I can't find the definition for the struct "jprobe" anyplace. Am I
missing some patch against my kernel?
Peter
|
|
From: Wolfram G. <Wol...@mi...> - 2005-04-08 12:25:20
|
Hi, my ARM port is making progress. Basic kprobes are running (pre_handler, s= ingle step, post_handler, fault_handler). What I have to do is the fixup of sin= gle stepped instructions.=20 Sparc64 is a machine without hardware signle step support as ARM. Is it c= orrect that you cannot single step an instruction which changes the program coun= ter? =20 Small example below shows the original code with an instruction which mov= es the control to the kprobe_handler. The kprobe_handler single steps the copy. = If ORIG_DO just do something then BKPT is hit and control goes back to kprob= es the post_handler is called and everything is fine. But if ORIG_DO branches to somewhere kprobes have no possibilty to recognize that and the post_handl= er is never called, correct? original code DO DO code stored in p->insn[] BKPT-> moves control to the kprobe_handler->ORIG_DO (->possibly branches = away) DO call post_handler an do fixup <----BKPT=20 DO The only solution I see is to set a BKPT after ORIG_DO and at the branch = target. But perhaps I am missing something. Thanks Wolfram --=20 Dipl.-Ing. Wolfram Gettert Software Engineer Embedded Systems Mixed Mode GmbH Lochhamer Schlag 17 D-82166 Gr=E4felfing/M=FCnchen Tel. +49-89-89868.0 Fax +49-89-89868.222 mailto:Wol...@mi... http://www.mixed-mode.de Mixed Mode - Your Partner for Systems Engineering! |
|
From: <p.l...@co...> - 2005-04-07 20:21:47
|
Looking further at the code, this seems to be a bug in the manner that dprobes and LTT are looking for trace configurations. The original way to turn on LTT was CONFIG_TRACE, but for some reason this has migrated to CONFIG_LTT. This older config option is not available. Hence, when dprobes was compiled, the code to interface with LTT was never compiled in. Also, some of the routine have changed names, from trace_<foo> to ltt_<foo>. I am attempting to change these and recompile the kernel, and see if that works. I will also log a bug with SuSE. In looking at the latest releases of each package, there seems to still be a difference with dprobes using CONFIG_TRACE and LTT using CONFIG_LTT. Is no one using these tools together? I am surprised if this has not been caught before. Regards, Peter > After struggling with LTT and dprobes on SLES9, I have gotten them to work > independently, but cannot get them to work together. I am probably overlooking > something obvious. [longwinded description deleted] |
|
From: Wolfram G. <Wol...@mi...> - 2005-04-04 14:15:49
|
Hi, in older implementations of the kprobes I have found arch_prepare_kprobe, arch_copy_kprobe, arch_remove_kprobe functions. Only arch_prepare_kprobe() is called from kernel/kprobes.c now. I haven't found any replacement for arch_remove_kpr= obe(). Have I missed something? Wolfram --=20 Dipl.-Ing. Wolfram Gettert Software Engineer Embedded Systems Mixed Mode GmbH Lochhamer Schlag 17 D-82166 Gr=E4felfing/M=FCnchen Tel. +49-89-89868.0 Fax +49-89-89868.222 mailto:Wol...@mi... http://www.mixed-mode.de Mixed Mode - Your Partner for Systems Engineering! |
|
From: Tom Z. <za...@us...> - 2005-04-04 07:20:16
|
Peter Lauterbach writes: > I have been trying to get dprobes working to trace some kernel functions > with LTT. > > I am using SuSE 9.2 Professional, but it does not seem to have the LTT > patches against the kernel. So I pulled a kernel from SLES9, and > compiled with all the appropriate switches in the .config and changes > dprobes requires per the README.dpcc. I can take an LTT trace with the > default trace points. Since it uses dprobes under the hood, it is safe > to assume dprobes is working in some fashion? Hi, LTT doesn't use dprobes under the hood, so you can't assume anything about dprobes from the fact that LTT is working. > > Compiling and running with the SLES9 kernel on SuSE 9.2 seems OK, but > when I compile something like tut14.dpc or the examples out of the > dprobes.lang man page, I just get a segmentation violation when it reads > the symbol table in /usr/src/linux/vmlinux. There is a known problem with dpcc in SLES9. I have a patch that was supposed to go into SLES9 SP1 to fix that, but it was somehow lost in the shuffle and never applied. I'll send that to you tomorrow. It sounds though that you're getting a segmentation violation with the dprobes command itself - I haven't seen that, but it sounds like a misconfiguration if so. In any case, I'll be happy to help you get it all working and your custom events out of dprobes and into LTT - I know it all works given the proper setup, as I've done the same thing myself. I'll send more info along with the dpcc patch tomorrow... Tom |
|
From: Prasanna S P. <pra...@in...> - 2005-04-04 06:21:58
|
Hi Peter, > > I have been trying to get dprobes working to trace some kernel functions > with LTT. > > I am using SuSE 9.2 Professional, but it does not seem to have the LTT > patches against the kernel. So I pulled a kernel from SLES9, and > compiled with all the appropriate switches in the .config and changes > dprobes requires per the README.dpcc. I can take an LTT trace with the > default trace points. Since it uses dprobes under the hood, it is safe > to assume dprobes is working in some fashion? > > Compiling and running with the SLES9 kernel on SuSE 9.2 seems OK, but > when I compile something like tut14.dpc or the examples out of the > dprobes.lang man page, I just get a segmentation violation when it reads > the symbol table in /usr/src/linux/vmlinux. > > dprobes itself seems OK, "dprobes -q -a -x" returns "No information > returned." as it should. > > Please advise what to check to get dprobes working with custom probes. I > really want to do what is close to the long example 6 back in dprobes.lang. > > Is there a beginners guide to writeing and troubleshooting rpn files? > Most of the stuff I have found is a high level overview of dprobes > architecture, or really detailed advanced features, not much in between. > > Regards, > Peter > The man pages will help you a lot to understand and use dprobes. Can you tell what command (with options)did you use to insert probe that failed. If dprobes is used with correct options it works fine. Thanks Prasanna -- Prasanna S Panchamukhi Linux Technology Center India Software Labs, IBM Bangalore Ph: 91-80-25044636 <pra...@in...> |
|
From: Peter L. <p.l...@co...> - 2005-04-04 01:35:59
|
I have been trying to get dprobes working to trace some kernel functions with LTT. I am using SuSE 9.2 Professional, but it does not seem to have the LTT patches against the kernel. So I pulled a kernel from SLES9, and compiled with all the appropriate switches in the .config and changes dprobes requires per the README.dpcc. I can take an LTT trace with the default trace points. Since it uses dprobes under the hood, it is safe to assume dprobes is working in some fashion? Compiling and running with the SLES9 kernel on SuSE 9.2 seems OK, but when I compile something like tut14.dpc or the examples out of the dprobes.lang man page, I just get a segmentation violation when it reads the symbol table in /usr/src/linux/vmlinux. dprobes itself seems OK, "dprobes -q -a -x" returns "No information returned." as it should. Please advise what to check to get dprobes working with custom probes. I really want to do what is close to the long example 6 back in dprobes.lang. Is there a beginners guide to writeing and troubleshooting rpn files? Most of the stuff I have found is a high level overview of dprobes architecture, or really detailed advanced features, not much in between. Regards, Peter |
|
From: jaimin.bhavsar <jai...@in...> - 2005-03-03 11:42:59
|
Hi , I was executing testcases on kernel modules from dprobes-testsuite-version1.1.0. On Redhat 9.0 system running Linux kernel 2.6.2 on i386 system. All the "kernel module" test cases which are part of dprobe testsuite version 1.1.0 failed. When I tried manually compiling ".dpc" file. as below : dpcc -I dpcs/include dpcs/module/builtin-outw.dpc -o builtin-outw.rpn I got following error message : loading symbol file : dprobes-testsuite-ver1.1.0/trigger-program/module/dprobe-testsuite-module/triggerprobe.ko (no debugging symbol found) unable to find function : test_fn I was not able to compile the above .dpc file. Also the perl script which runs tests in automated fashion fails compiling the test cases, and all the test cases fails. I verfied using "nm" command that, "test_fn" symbol exists in the triggerprobe.ko module. and "test_fn" is invoked when the module is inserted in to the kernel. But "dpcc" fails locating the "test_fn" symbol in triggerprobe.ko module. Kindly help me out in compiling this test cases. Thanks in advance Regards Jaimin |
|
From: <ev...@te...> - 2005-02-18 21:47:33
|
Karim, I appreciate your viewpoint. It would be interesting to see if IBM's legal resources have an opinion. I haven't had a chance to look at the source or the license yet but I had used DTrace when I worked at sun and it was tres nifty. -evt ---- Original message ---- >Date: Fri, 18 Feb 2005 13:20:06 -0500 >From: Karim Yaghmour <ka...@op...> >Subject: Re: [Dprobes] has anyone scoped the effort to port DTrace >To: ev...@te... >Cc: dp...@ww... > > >ev...@te... wrote: >> now that sun has released the code it's worth a look. > >To the best of my understanding the licenses are incompatible ... >intentionally. Like in other cases, looking at the DTrace code would >probably be a trap for any open source developer. > >Karim >-- >Author, Speaker, Developer, Consultant >Pushing Embedded and Real-Time Linux Systems Beyond the Limits >http://www.opersys.com || ka...@op... || 1-866-677-4546 |
|
From: <ev...@te...> - 2005-02-18 20:22:45
|
now that sun has released the code it's worth a look. -evt |
|
From: Andi K. <ak...@su...> - 2005-02-18 13:13:33
|
On Fri, Feb 18, 2005 at 02:50:09PM +0530, Nilesh Hiray wrote: > Hi all, > We have a Suse 9 Linux (SUSE LINUX Enterprise Server 9 for x86) > installed. It looks like it comes with DProbes fully installed on it. > However when I use /DProbes -i test.probe/ it fails saying /"DProbes > device not found". / > Do I have to install, the Dprobes module (I guess what is called the > "non-kernel portion") manually or it gets automatically installed with > Suse ? > If yes, where do i get the source rpm, the Suse had version 3.6.5 of > Dprobes which is not available anywhere online ? dprobes is currently only enabled in the "kernel-debug" kernel rpm. You would need to run that. -Andi |
16 messages has been excluded from this view by a project administrator.
