* [Bug uprobes/6879] New: does not currently support probing instruction first byte 0x65
@ 2008-09-11 21:20 fche at redhat dot com
2008-09-11 23:24 ` [Bug uprobes/6879] " jkenisto at us dot ibm dot com
` (4 more replies)
0 siblings, 5 replies; 6+ messages in thread
From: fche at redhat dot com @ 2008-09-11 21:20 UTC (permalink / raw)
To: systemtap
A para-callgraph.stp trace of libc produces several dozen such printk warnings
per run. This is on i686 rhel5.2ish, /lib/libc-2.5.so, __res_init, note
address 0x72e32a:
0072e310 <__res_init>:
72e310: 55 push %ebp
72e311: 89 e5 mov %esp,%ebp
72e313: 83 ec 14 sub $0x14,%esp
72e316: 89 5d f4 mov %ebx,0xfffffff4(%ebp)
72e319: e8 12 49 f3 ff call 662c30 <__i686.get_pc_thunk.bx>
72e31e: 81 c3 d6 dc 05 00 add $0x5dcd6,%ebx
72e324: 89 75 f8 mov %esi,0xfffffff8(%ebp)
72e327: 89 7d fc mov %edi,0xfffffffc(%ebp)
72e32a: 65 a1 00 00 00 00 mov %gs:0x0,%eax
72e330: 03 83 bc fe ff ff add 0xfffffebc(%ebx),%eax
72e336: 8b 00 mov (%eax),%eax
72e338: 8b 30 mov (%eax),%esi
--
Summary: does not currently support probing instruction first
byte 0x65
Product: systemtap
Version: unspecified
Status: NEW
Severity: normal
Priority: P2
Component: uprobes
AssignedTo: systemtap at sources dot redhat dot com
ReportedBy: fche at redhat dot com
GCC host triplet: i686
http://sourceware.org/bugzilla/show_bug.cgi?id=6879
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Bug uprobes/6879] does not currently support probing instruction first byte 0x65
2008-09-11 21:20 [Bug uprobes/6879] New: does not currently support probing instruction first byte 0x65 fche at redhat dot com
@ 2008-09-11 23:24 ` jkenisto at us dot ibm dot com
2008-09-11 23:25 ` jkenisto at us dot ibm dot com
` (3 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: jkenisto at us dot ibm dot com @ 2008-09-11 23:24 UTC (permalink / raw)
To: systemtap
------- Additional Comments From jkenisto at us dot ibm dot com 2008-09-11 23:23 -------
Created an attachment (id=2944)
--> (http://sourceware.org/bugzilla/attachment.cgi?id=2944&action=view)
Enable probing of instructions with the 0x65 (gs:) prefix
Here's a patch that changes i386, x86_64, and x86 uprobes so that they don't
reject probes on instructions with the 0x65 prefix. I haven't tested it.
Frank, could you give it a try?
I scanned a disassembly of libc on my i686 system and verified that none of the
other segment prefixes are used, but I haven't done an exhaustive search for
all possible opcodes.
--
http://sourceware.org/bugzilla/show_bug.cgi?id=6879
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Bug uprobes/6879] does not currently support probing instruction first byte 0x65
2008-09-11 21:20 [Bug uprobes/6879] New: does not currently support probing instruction first byte 0x65 fche at redhat dot com
2008-09-11 23:24 ` [Bug uprobes/6879] " jkenisto at us dot ibm dot com
@ 2008-09-11 23:25 ` jkenisto at us dot ibm dot com
2008-09-12 3:53 ` fche at redhat dot com
` (2 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: jkenisto at us dot ibm dot com @ 2008-09-11 23:25 UTC (permalink / raw)
To: systemtap
--
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |WAITING
http://sourceware.org/bugzilla/show_bug.cgi?id=6879
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Bug uprobes/6879] does not currently support probing instruction first byte 0x65
2008-09-11 21:20 [Bug uprobes/6879] New: does not currently support probing instruction first byte 0x65 fche at redhat dot com
2008-09-11 23:24 ` [Bug uprobes/6879] " jkenisto at us dot ibm dot com
2008-09-11 23:25 ` jkenisto at us dot ibm dot com
@ 2008-09-12 3:53 ` fche at redhat dot com
2008-09-17 16:14 ` fche at redhat dot com
2008-11-18 23:35 ` jkenisto at us dot ibm dot com
4 siblings, 0 replies; 6+ messages in thread
From: fche at redhat dot com @ 2008-09-12 3:53 UTC (permalink / raw)
To: systemtap
------- Additional Comments From fche at redhat dot com 2008-09-12 03:52 -------
The new code makes
probe process("/lib/libc-2.5.so").function("*").call { }
probes work well enough to insert them all. It turns out we
later get
"Task died at uprobe probepoint: ... probepoint = 0xaddr"
messages which I'm tracking down.
As for x86-64, we need first-byte-0x64 treatment:
0000003c2ea990a0 <__getpid>:
3c2ea990a0: 64 8b 14 25 94 00 00 mov %fs:0x94,%edx
3c2ea990a7: 00
3c2ea990a8: 83 fa 00 cmp $0x0,%edx
3c2ea990ab: 89 d0 mov %edx,%eax
--
What |Removed |Added
----------------------------------------------------------------------------
Status|WAITING |NEW
Last reconfirmed|0000-00-00 00:00:00 |2008-09-12 03:52:10
date| |
http://sourceware.org/bugzilla/show_bug.cgi?id=6879
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Bug uprobes/6879] does not currently support probing instruction first byte 0x65
2008-09-11 21:20 [Bug uprobes/6879] New: does not currently support probing instruction first byte 0x65 fche at redhat dot com
` (2 preceding siblings ...)
2008-09-12 3:53 ` fche at redhat dot com
@ 2008-09-17 16:14 ` fche at redhat dot com
2008-11-18 23:35 ` jkenisto at us dot ibm dot com
4 siblings, 0 replies; 6+ messages in thread
From: fche at redhat dot com @ 2008-09-17 16:14 UTC (permalink / raw)
To: systemtap
------- Additional Comments From fche at redhat dot com 2008-09-17 16:12 -------
Jim's patch 85d2b7f committed.
--
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |FIXED
http://sourceware.org/bugzilla/show_bug.cgi?id=6879
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
^ permalink raw reply [flat|nested] 6+ messages in thread
* [Bug uprobes/6879] does not currently support probing instruction first byte 0x65
2008-09-11 21:20 [Bug uprobes/6879] New: does not currently support probing instruction first byte 0x65 fche at redhat dot com
` (3 preceding siblings ...)
2008-09-17 16:14 ` fche at redhat dot com
@ 2008-11-18 23:35 ` jkenisto at us dot ibm dot com
4 siblings, 0 replies; 6+ messages in thread
From: jkenisto at us dot ibm dot com @ 2008-11-18 23:35 UTC (permalink / raw)
To: systemtap
------- Additional Comments From jkenisto at us dot ibm dot com 2008-11-18 23:34 -------
Applied same fixes to runtime/uprobes2 -- commit 8590d03.
--
http://sourceware.org/bugzilla/show_bug.cgi?id=6879
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2008-11-18 23:35 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-09-11 21:20 [Bug uprobes/6879] New: does not currently support probing instruction first byte 0x65 fche at redhat dot com
2008-09-11 23:24 ` [Bug uprobes/6879] " jkenisto at us dot ibm dot com
2008-09-11 23:25 ` jkenisto at us dot ibm dot com
2008-09-12 3:53 ` fche at redhat dot com
2008-09-17 16:14 ` fche at redhat dot com
2008-11-18 23:35 ` jkenisto at us dot ibm dot com
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).