Commit Graph

62 Commits

Author SHA1 Message Date
Dag-Erling Smørgrav
95bb676aee Address some of bde@'s concerns with the new code. 2011-10-08 21:02:08 +00:00
Dag-Erling Smørgrav
515dfc239d Teach kdump(1) to decode capability bitmasks.
MFC after:	3 weeks
2011-10-08 13:01:38 +00:00
Dag-Erling Smørgrav
d09e66be12 Bring ioctlname() in line with all the other *name() functions, which
actually print the name (or the numeric value, if they can't figure out
the correct name) instead of just returning a pointer to it.  Also, since
ioctl numbers are not and probably never will be unique, drop support for
using a switch statement instead of an if/else chain.
2011-10-08 12:47:00 +00:00
Dag-Erling Smørgrav
09307b2547 I appreciate the logic behind using a (void) cast to indicate that the
return value is intentionally ignored, but frankly, all it does is
get in the way of the code.

Also fix a few other incorrect casts, such as (void *)malloc(foo) and
passing signed values to %x.
2011-10-08 12:28:06 +00:00
Dag-Erling Smørgrav
e4bb45b1ce Fix casting. 2011-10-08 12:21:51 +00:00
Dag-Erling Smørgrav
c02541ee80 Whitespace. 2011-10-08 12:10:16 +00:00
Dag-Erling Smørgrav
73430055e6 C has had swicth statements for 40 years or so. It's about time we
started using them.
2011-10-08 11:51:48 +00:00
Dag-Erling Smørgrav
34bb8e6d12 1) Some of the #defines or enums for which we auto-generate naming
functions may be wider than int, so use intmax_t throughout.  Also
   add missing casts in printf() calls.

2) Clean up some of the auto-generated code to improve readability.

3) Auto-generate kdump_subr.h.  Note that this requires a semi-ugly hack
   in the Makefile to make sure it is generated before make(1) tries to
   build kdump.c, or preprocess it for 'make depend'.

MFC after:	3 weeks
2011-10-08 09:57:29 +00:00
Jilles Tjoelker
a4a1c5b11e kdump: Show code for signals where the default action was taken.
This information is available as of kernel r220740. Trace files from older
kernels will always have 0.

MFC after:	1 week
2011-04-17 21:11:55 +00:00
Dmitry Chagin
dde5f9b938 Teach kdump to decode linux syscalls names too.
Fix bug introduced in my previous commit: the kernel always dump native
signal numbers, so no need to check the ABI in ktrpsig().

Suggested by:	jhb
MFC after:	1 Month.
2011-03-01 16:42:28 +00:00
Dmitry Chagin
7144d36fab Teach kdump to understand sv_flags records in the trace files.
MFC after:	1 Month.
2011-02-25 22:07:23 +00:00
Joel Dahl
da52b4caaf Remove the advertising clause from UCB copyrighted files in usr.bin. This
is in accordance with the information provided at
ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change

Also add $FreeBSD$ to a few files to keep svn happy.

Discussed with:	imp, rwatson
2010-12-11 08:32:16 +00:00
Ed Schouten
99742a231f Change all our own code to use st_*tim instead of st_*timespec.
Also remove some local patches to diff(1) which are now unneeded.
2010-03-28 13:16:08 +00:00
Warner Losh
30c1278f14 The kdump data stream is an unaligned data stream for stat and
sockaddr structures.  As such, we have top copy the data structure
into a local buffer before we can reference it, otherwise we have
unaligned references (these are fixed up automatically on some CPUs,
but not on others).  We do this unconditionally to make the code
easier to read and understand.

Submitted by:	Grzegorz Bernacki
2010-02-18 16:05:09 +00:00
Jaakko Heinonen
c065c0b8e1 - Cast intptr_t, pid_t and time_t values to intmax_t and use %jd with
printf.
- Cast the system call return value to long and use %ld in a printf in
  ktrsysret().

PR:		bin/123774
MFC after:	2 weeks
2010-02-06 16:01:38 +00:00
Colin Percival
22ce05c97c Update malloc utrace structure parsing to reflect the change (r199265)
in how malloc_init is logged from (0, 0, 0) to (-1, 0, 0).

While we're here, simplify the logic.

Reviewed by:	jhb (earlier version)
2009-11-14 09:33:51 +00:00
Attilio Rao
48f6dd8a8c Use a safety belt for cases where corrupted narg can be passed to the
ktrsyscall(). print_number() does decrement the number of arguments,
leading to infinite loops for negative values.

Reported by:	Patrick Lamaiziere <patpr at davenulle dot org>,
		Jonathan Pascal <jkpyvxmzsa at mailinator dot com>
Submitted by:	jh
PR:		bin/120055, kern/119564
MFC:		1 week
2009-11-07 21:46:34 +00:00
Xin LI
1a604cfa07 Add two missing include files and prototype for sockfamilyname()
which is generated by mksubr.
2009-03-20 18:29:52 +00:00
John Baldwin
a56be37e68 Add a new type of KTRACE record for sysctl(3) invocations. It uses the
internal sysctl_sysctl_name() handler to map the MIB array to a string
name and logs this name in the trace log.  This can be useful to see
exactly which sysctls a thread is invoking.

MFC after:	1 month
2009-03-11 21:48:36 +00:00
Robert Watson
e4372ceba0 Remove netatm from HEAD as it is not MPSAFE and relies on the now removed
NET_NEEDS_GIANT.  netatm has been disconnected from the build for ten
months in HEAD/RELENG_7.  Specifics:

- netatm include files
- netatm command line management tools
- libatm
- ATM parts in rescue and sysinstall
- sample configuration files and documents
- kernel support as a module or in NOTES
- netgraph wrapper nodes for netatm
- ctags data for netatm.
- netatm-specific device drivers.

MFC after:	3 weeks
Reviewed by:	bz
Discussed with:	bms, bz, harti
2008-05-25 22:11:40 +00:00
Ruslan Ermilov
da647ae946 Spell -t option's argument by name. 2008-04-02 09:41:29 +00:00
Jeff Roberson
0f33b907d5 - Remove ksethrcmdname. 2008-03-12 11:51:01 +00:00
Dag-Erling Smørgrav
60e15db992 This patch adds a new ktrace(2) record type, KTR_STRUCT, whose payload
consists of the null-terminated name and the contents of any structure
you wish to record.  A new ktrstruct() function constructs and emits a
KTR_STRUCT record.  It is accompanied by convenience macros for struct
stat and struct sockaddr.

In kdump(1), KTR_STRUCT records are handled by a dispatcher function
that runs stringent sanity checks on its contents before handing it
over to individual decoding funtions for each type of structure.
Currently supported structures are struct stat and struct sockaddr for
the AF_INET, AF_INET6 and AF_UNIX families; support for AF_APPLETALK
and AF_IPX is present but disabled, as I am unable to test it properly.

Since 's' was already taken, the letter 't' is used by ktrace(1) to
enable KTR_STRUCT trace points, and in kdump(1) to enable their
decoding.

Derived from patches by Andrew Li <andrew2.li@citi.com>.

PR:		kern/117836
MFC after:	3 weeks
2008-02-23 01:01:49 +00:00
John Baldwin
0460d3516d Only use sockoptname() to parse socket option names for SOL_SOCKET
requests.

MFC after:	3 days
Reported by:	Michiel Boland  michiel boland.org
2008-01-07 18:50:25 +00:00
John Baldwin
b850a685ce Add several missing comma's in the fancy syscall parsing case before
some arguments that are parsed (e.g., semctl command names).

MFC after:	3 days
2007-12-06 08:49:43 +00:00
John Baldwin
1db0049fa1 Fix alignment of context switch traces.
MFC after:	1 week
Approved by:	re (rwatson: "I like simple patches.")
2007-07-10 00:01:30 +00:00
Peter Wemm
cae1120adf kdump has knowledge of lseek() and mmap() arg decoding. Teach it about
the new mmap and lseek syscalls.

Approved by:  re (kensmith)
2007-07-04 23:32:49 +00:00
Ed Maste
24be4e6248 Remove static ptrace_ops array and extract ptrace op names from
sys/ptrace.h with mksubr.
2007-04-09 19:16:24 +00:00
John Baldwin
1f4b63f824 Add various utrace's for use with ktrace to the ELF runtime linker. To
activate the traces, set the LD_UTRACE (or LD_32_UTRACE) environment
variable.  This also includes code in kdump(8) to parse the traces.

Reviewed by:	kan, jdp
MFC after:	2 weeks
2007-01-09 17:50:05 +00:00
John Baldwin
670b9e9fda Add code to parse the utrace(2) entries generated by malloc(3) in a more
human-readable format.  Note that we report 'realloc(p, 0)' as 'free(p)'
since both cases are encoded the same way and 'free()' is more common
than a realloc() to 0.

MFC after:	1 week
2007-01-05 21:04:37 +00:00
Craig Rodrigues
8bc31d836f Add sockipprotoname() function. Decode the third parameter (protocol)
of a socket() call with sockipprotoname() if the first parameter (domain)
is PF_INET or PF_INET6.

Old parsing behavior before this change:
ping6    CALL  socket(PF_INET6,SOCK_RAW,0x3a)

New behavior after this change:
ping6    CALL  socket(PF_INET6,SOCK_RAW,IPPROTO_ICMPV6)
2007-01-04 04:46:59 +00:00
Maxim Konovalov
5a3e0a1b99 o Fix some printf(3) format nits in my patch I submitted to kib@.
Submitted by:	ru
2006-07-12 13:36:08 +00:00
Konstantin Belousov
fd6afe791c Check that the signal number is in range.
Submitted by:	maxim
MFC after:	1 week
Approved by:	pjd (mentor)
2006-07-12 12:41:56 +00:00
Alexander Leidinger
98a68a586f Change kdump to print more useful information, i.e. it changes from
32229 telnet   CALL  mmap(0,0x8000,0x3,0x1002,0xffffffff,0,0,0)
 32229 telnet   CALL  open(0x2807bc28,0,0x1b6)
 32229 telnet   CALL  socket(0x2,0x2,0)
to
 32229 telnet   CALL mmap(0,0x8000,PROT_READ|PROT_WRITE,MAP_PRIVATE|MAP_ANON,0xffffffff,0,0,0)
 32229 telnet   CALL  open(0x2807bc28,O_RDONLY,<unused>0x1b6)
 32229 telnet   CALL  socket(PF_INET,SOCK_DGRAM,0)

David wanted to implement the suggestions which came up at the review from
arch@ too, but real life rejected this proposal. So I commit what we already
got and let another volunteer pick the remaining work from the ideas list.

Submitted by:	"David Kirchner" <dpk@dpk.net>
Suggested by:	FreeBSD ideas list page
Reviewed by:	arch
2006-05-20 14:27:22 +00:00
Robert Watson
df0c8868b1 Add "-s" argument to kdump to suppress the display of I/O data.
MFC after:	1 week
2005-11-12 14:21:48 +00:00
Robert Watson
a9ac598b36 Add a "-H" argument to kdump, which causes kdump to print an additional
field holding the threadid.  This is more useful for libthr than
libpthread, but still quite useful in libpthread as it can be used to
process interlaced records from multiple threads over the course of a
system call.

Detect old ktr_buffer values using the heuristic "if it's negative,
then it must not be a valid threadid".  This may leave something to be
desired.

MFC after:	1 month
Reviewed by:	davidxu
2005-11-01 14:48:23 +00:00
Ruslan Ermilov
7143dfdc43 Place a function prototype correctly.
Submitted by:	Divacky Roman
2004-09-19 14:54:35 +00:00
Poul-Henning Kamp
743f91743d Respect decimal flag when dumping USER type records. 2004-03-25 12:33:55 +00:00
Peter Wemm
24b648094d First pass at attempted debrucification:
- sort the -E switch into the right place.
- add previously missing -p pid in usage (from the last few commits).
- add -E to usage.
- consistently use trfile in the man page.

I knew I shouldn't have touched the man page.  If I commit to a man page,
it just makes people suspicious. :-)
2003-12-08 22:47:59 +00:00
Peter Wemm
ec13191464 Add a -E (elapsed time) flag to kdump. This is like -T, except it is
relative to start of the dump.

Approved by:  re (scottl)
2003-12-07 01:06:32 +00:00
Peter Wemm
ec4beb5d8b Two enhancements for kdump.
1) add a "-p pid", which is rather useful for selecting a single pid in
a combined trace file (eg: with ktrace -i).
2) display binary genio data in a more precise format.
2003-06-03 01:44:43 +00:00
John Baldwin
b9609ab177 Add a default case that just outputs a new line for the case of an
unknown header type.
2003-03-13 18:58:39 +00:00
John Baldwin
db53f66b42 Teach kdump(8) to handle events marked with KTR_DROP. If a record has
KTR_DROP set in its header, then we output an extra line to stdout to
indicate that events were dropped between the previous record and this
record.  It is a bit trickier because we need to always notify the user
if events are dropped even if KTR_DROP is set on a record of a type that
we aren't interested in since kdump(8) doesn't know if the dropped events
were of the types that the user has requested.  To avoid outputting
multiple events dropped notices in between actual event logs, a state
variable is set whenever a drop is logged and cleared whenever an actual
record is output.

Requested by:	phk
2003-03-13 18:46:35 +00:00
David Malone
91fbb9c1db Improve WARNS situation for kdump:
1) Define _KERNEL while including sys/time.h to get some function prototypes.
2) Add prototypes and ANSIify definitions.
3) Constness changes.
4) Remove register keyword.
5) Actually return a sensible value from main.
6) Make fread_tail take a void * instead of a char *.
7) Avoid a signedness warning by casting to a size_t. Should be safe
   enough 'cos we also check for nonnegativity.
8) Be extra chummy with sigset_t rather than passing a struct to printf
   and pretending it is an int.
2002-07-28 16:05:07 +00:00
David E. O'Brien
e026a48c34 Consistently use FBSDID 2002-06-30 05:25:07 +00:00
Peter Wemm
c447342094 Change #ifdef KERNEL to #ifdef _KERNEL in the public headers. "KERNEL"
is an application space macro and the applications are supposed to be free
to use it as they please (but cannot).  This is consistant with the other
BSD's who made this change quite some time ago.  More commits to come.
1999-12-29 05:07:58 +00:00
Peter Wemm
c3aac50f28 $Id$ -> $FreeBSD$ 1999-08-28 01:08:13 +00:00
Bruce Evans
c52c4cb57d Decode new ptrace requests PT_GETDBREGS and PT_SETDBREGS. 1999-07-17 09:27:41 +00:00
Bruce Evans
9810672047 Decode all currently supported values of the ptrace `request' arg. 1999-06-26 07:31:13 +00:00
Dmitrij Tejblum
7a6ec4cfb2 Syscall arguments are now properly aligned. Print them, and syscall return
values, as longs, instead of int.
1999-06-16 18:46:40 +00:00