Commit Graph

106 Commits

Author SHA1 Message Date
Alfred Perlstein
d10f73b311 don't seg if the syscall is invalid for i386fbsd. 2004-08-08 23:29:36 +00:00
Stefan Farfeleder
14cc87e479 Use the length modifier 'll' instead of 'q' to print long longs. 2004-07-28 16:03:13 +00:00
Alfred Perlstein
2bae4eb308 Support readlink(2) better. Readlink does not nul terminate the
result buffer, so we need to format it ourselves.  The problem is
that the length is stored as the return value from readlink, so we
need to pass the return value from our syscall into print_arg.

Motivated by: truss garbage on my screen from reading /etc/malloc.conf.
2004-07-17 19:48:49 +00:00
Alfred Perlstein
d75300bfa5 When reporting reciept of a signal, print the signal's name. 2004-07-17 19:19:36 +00:00
Alfred Perlstein
fde3a7d1e9 Decode the "wence" arg to lseek and linux_lseek. 2004-07-10 09:23:53 +00:00
Bruce M Simpson
63fb96902c Fix unterminated RCSID.
Submitted by:	Liam J. Foy
2004-06-13 18:06:49 +00:00
David Malone
a5d31d1631 A first stab at truss support for amd64, basically cogged from i386.
It seems to work in my limited tests.
2004-06-12 22:49:06 +00:00
Poul-Henning Kamp
01b8b4dc80 COMPAT_SUNOS is gone. 2004-06-11 11:58:07 +00:00
David Malone
e45a5a0d08 Add support for decoding Timespec, Timeval, Itimerval, Pollfd,
Fd_set and Sigaction structures. Use these for printing the arguments
to sigaction(), nanosleep(), select(), poll(), gettimeofday(),
clock_gettime(), recvfrom(), getitimer() and setitimer().

This is based on Dan's patch from the PR but I've hacked it for
style and some other issues. While Dan has checked this patch, any
goofs are probably my fault.

(The PR also contains support for the dual return values of pipe().
These will follow once I've ported that support to platforms other
than i386.)

PR:		52190
Submitted by:	Dan Nelson <dnelson@allantgroup.com>
2004-03-23 12:37:02 +00:00
David Malone
b637237cba Use pread to implement pread, rather than taking a detour throug stdio.
PR:		52190
Submitted by:	Dan Nelson <dnelson@allantgroup.com>
2004-03-23 12:29:17 +00:00
Alfred Perlstein
5214320f76 decode mprotect args while i'm here. 2004-03-23 09:16:18 +00:00
Alfred Perlstein
894b8f7afb decode fcntl and mmap arguments. 2004-03-23 09:04:06 +00:00
David Malone
1bc99f1144 Move declarations of Procfd to a header file. 2004-01-07 14:29:45 +00:00
David Malone
9b83b9f75c Fix a printf format warning. 2004-01-07 14:27:30 +00:00
Martin Cracauer
5cdf6a132b Fix signal behaviour.
In my last change I made sure that the signal as reported from a truss
exit is the same as if truss wasn't between parent and trussed
program.  I was smart enough to not have it coredump on SIGQUIT but it
didn't ocur to me SIGSEGV might cause a coredump, too :-)

So get rid of SIGQUIT extra hack and limit coredumpsize to zero
instead.

Tested: still works, correct signal reported.  No more codedumps from
SIGSEGV in the trussed proces.  This file compiles cleanly on AMD64
(sledge).

PR:
Submitted by:
Reviewed by:
Approved by:
Obtained from:
MFC after:
2003-12-28 01:20:03 +00:00
Marcel Moolenaar
4e92419dcd Do not ignore any possible errors that fseeko() may have. The fact
is that fseeko() fails in very predictable and frequent ways on ia64.
This is because the offset is actually an address in the process'
address space, which on ia64 can be larger than long (for lseek) or
off_t (for fseeko). The crux is the signedness. The register stack
and memory stack are in region 4 on ia64. This means that the sign bit
is 1. The large positive virtual address is wrongly interpreted as
a negative file offset.

There's no quick fix. Even if you get around the API by using a
SEEK_SET up to LONG_MAX and follow it up with a SEEK_CUR for the
remainder, the kernel simply cannot deal with it. and the second
seek will just fail.

Therefore, this change does not actually fix the root cause. It just
makes sure we're not spitting out all kinds of garbage or that the
get_struct() function in particular does not cause truss(1) to exit.
This, I might add, invariably happened way too soon for truss(1) to
be of any use on ia64...
2003-11-13 09:04:24 +00:00
Marcel Moolenaar
1bcb5f5a96 Port truss(1) to 64-bit architectures:
o  Syscall return values do not fit in int on 64-bit architectures.
   Change the type of retval in <arch>_syscall_exit() to long and
   change the prototype of said function to return a long as well.
o  Change the prototype of print_syscall_ret() to take a long for
   the return address and change the format string accordingly.
o  Replace the code sequence
	tmp = malloc(X);
	sprintf(tmp, format, ...);
   with X by definition too small on 64-bit platforms by
        asprintf(&tmp, format, ...);

With these changes the output makes sense again, although it does
mess up the tabulation on ia64. Go widescreen...

Not tested on: alpha, sparc64.
2003-11-09 03:48:13 +00:00
Marcel Moolenaar
a7a08c7e63 Fix truss so that it doesn't abort/exit when a syscall has been given
a NULL-pointer for a sockaddr argument.
2003-10-27 06:50:57 +00:00
Philippe Charnier
b956c13cfe Add FBSDID. Do not \n terminate warnx() argument. fprint() -> warnx(). 2003-09-07 15:50:43 +00:00
Dag-Erling Smørgrav
99d1300607 send() and recv() are just wrappers, not actual syscalls. 2003-08-05 19:34:36 +00:00
Dag-Erling Smørgrav
9b2f78c553 Add recv(2), recvfrom(2), send(2) and sendfrom(2). 2003-08-05 10:45:06 +00:00
Marcel Moolenaar
a9fdd3a89e Fix truss on ia64. The syscall arguments are written to the trap
frame, occupying scratch registers r16 and up. We don't have to
save any scratch registers for syscalls, so we have plenty of
room there. Consequently, when we fetch the registers from the
process, we automaticly have all the arguments and don't need
to read them seperately.
2003-08-02 22:29:10 +00:00
Marcel Moolenaar
1eb3be492d Use cat(1) instead of cp(1) so as not to break -DNOCLEAN builds
when the file permissions of source files don't allow writing.
2003-06-06 05:38:09 +00:00
Marcel Moolenaar
f2c49dd248 Revamp of the syscall path, exception and context handling. The
prime objectives are:
o  Implement a syscall path based on the epc inststruction (see
   sys/ia64/ia64/syscall.s).
o  Revisit the places were we need to save and restore registers
   and define those contexts in terms of the register sets (see
   sys/ia64/include/_regset.h).

Secundairy objectives:
o  Remove the requirement to use contigmalloc for kernel stacks.
o  Better handling of the high FP registers for SMP systems.
o  Switch to the new cpu_switch() and cpu_throw() semantics.
o  Add a good unwinder to reconstruct contexts for the rare
   cases we need to (see sys/contrib/ia64/libuwx)

Many files are affected by this change. Functionally it boils
down to:
o  The EPC syscall doesn't preserve registers it does not need
   to preserve and places the arguments differently on the stack.
   This affects libc and truss.
o  The address of the kernel page directory (kptdir) had to
   be unstaticized for use by the nested TLB fault handler.
   The name has been changed to ia64_kptdir to avoid conflicts.
   The renaming affects libkvm.
o  The trapframe only contains the special registers and the
   scratch registers. For syscalls using the EPC syscall path
   no scratch registers are saved. This affects all places where
   the trapframe is accessed. Most notably the unaligned access
   handler, the signal delivery code and the debugger.
o  Context switching only partly saves the special registers
   and the preserved registers. This affects cpu_switch() and
   triggered the move to the new semantics, which additionally
   affects cpu_throw().
o  The high FP registers are either in the PCB or on some
   CPU. context switching for them is done lazily. This affects
   trap().
o  The mcontext has room for all registers, but not all of them
   have to be defined in all cases. This mostly affects signal
   delivery code now. The *context syscalls are as of yet still
   unimplemented.

Many details went into the removal of the requirement to use
contigmalloc for kernel stacks. The details are mostly CPU
specific and limited to exception_save() and exception_restore().
The few places where we create, destroy or switch stacks were
mostly simplified by not having to construct physical addresses
and additionally saving the virtual addresses for later use.

Besides more efficient context saving and restoring, which of
course yields a noticable speedup, this also fixes the dreaded
SMP bootup problem as a side-effect. The details of which are
still not fully understood.

This change includes all the necessary backward compatibility
code to have it handle older userland binaries that use the
break instruction for syscalls. Support for break-based syscalls
has been pessimized in favor of a clean implementation. Due to
the overall better performance of the kernel, this will still
be notived as an improvement if it's noticed at all.

Approved by: re@ (jhb)
2003-05-16 21:26:42 +00:00
Matthew N. Dodd
726c2ca080 Decode a few more syscalls. 2003-04-15 06:12:12 +00:00
Ruslan Ermilov
568ef33fa2 Removed extra parentheses. 2003-02-20 15:09:52 +00:00
Ruslan Ermilov
66c8239c89 Fixed comment. 2003-02-20 15:05:39 +00:00
Ruslan Ermilov
06e482e60a mdoc(7) police: markup polishing.
Approved by:	re
2002-11-26 17:33:37 +00:00
Marcel Moolenaar
a3e32192a5 Port to ia64. It builds, but usability is very limited. 2002-11-10 00:59:13 +00:00
David Malone
eb9741e47d Various cleanups of truss:
1) Missing include.
2) Constness.
3) ANSIfication.
4) Avoid some shadowing.
5) Add/clarify some error messages.
6) Some int functions were using return without a value.
7) Mark some parameters as unused.
8) Cast a value we know is non-negative to a size_t before comparing.
2002-11-09 22:46:51 +00:00
David Malone
967706f308 Check if name is NULL before we strcmp the syscall name. This was
sometimes triggered if you began trussing a process while it was
in a syscall, as the name is filled in when the syscall is made.
2002-11-09 22:28:38 +00:00
Mike Barcroft
1d631f7ef1 Don't depend on <signal.h> to include <sys/time.h>, instead include
<sys/time.h> directly.  This is mostly bogus since CLOCK_REALTIME
should be defined in <time.h>, which these files already include.
2002-10-06 21:46:04 +00:00
Matthew N. Dodd
fb034d05ac Remove an unnecessary cast. 2002-08-07 11:35:18 +00:00
Matthew N. Dodd
37169f9446 - Use time.h not sys/time.h.
- Fix printf format errors.

Submitted by:	 bde
2002-08-06 12:46:14 +00:00
Matthew N. Dodd
28c4ea6da8 Remove extra space in output. 2002-08-05 14:56:20 +00:00
Matthew N. Dodd
42a389ac3e Diff reduction. 2002-08-05 13:23:41 +00:00
Matthew N. Dodd
203098d816 Use timespec not timeval. 2002-08-05 12:22:55 +00:00
Jake Burkholder
f84c971a6f Ported to sparc64. 2002-08-04 17:57:01 +00:00
Bruce Evans
c69d844294 Include <sys/time.h> for the declaration of struct timeval. Do not
depend on namespace pollution in <signal.h>.  (truss shouldn't be
using timevals anyway, since it was implemented long after timevals
were obsoleted by timespecs.)
2002-08-04 10:57:41 +00:00
Matthew N. Dodd
9897b20356 Add options to print the argument and environment string parameters to
execve().

This could be done in a more general manner but it still wouldn't
be very pretty.

MFC after:	 3 weeks
2002-08-04 02:24:21 +00:00
Matthew N. Dodd
0629483c41 Add a "FreeBSD ELF32" entry to ex_types[]. 2002-08-04 02:20:06 +00:00
Matthew N. Dodd
88a6987d40 Terminate the output line when a non-returning syscall is printed. 2002-08-04 01:28:13 +00:00
Matthew N. Dodd
0d0bd00edd Add options to print absolute and relative timestamps.
PR:		 bin/25587 (in part)
MFC after:	 3 weeks
2002-08-04 01:27:31 +00:00
Matthew N. Dodd
c03bfcc871 Allow tracking fork()ed children.
PR:		 bin/25587 (in part)
MFC after:	3 weeks
2002-08-04 01:02:52 +00:00
Matthew N. Dodd
ec0bed25ba Parameterize globals.
PR:		bin/25587 (in part)
MFC after:	3 weeks
2002-08-04 00:46:48 +00:00
Mark Murray
bfc3d86a24 s/inline/__inline/ 2002-07-19 13:49:37 +00:00
Peter Wemm
f81e3ffe10 OOPS! rev 1.16 accidently changed the default outfile from stderr to
stdout.  Unfortunately, DES mfc'ed this change in 1.15.2.1 (this
part probably should not have been) so it is broken there too.
truss is documented to use stderr, and other implementations use stderr.

Submitted by:	Arne Dag Fidjestøl <adf@idi.ntnu.no>
2002-05-16 21:58:57 +00:00
Peter Wemm
02b8e10aad Do not bother with #include <machine/psl.h> since it is just a stub file
that says something like "/* Not used on Alpha */".
2002-05-01 06:23:48 +00:00
David Malone
7c8225cab8 Use fseeko and uintptr_t to make sure that we get a sensible offset
when trying to read from the stack.

PR:		37104
Submitted by:	Thomas Quinot <thomas@cuivre.fr.eu.org>
MFC after:	3 weeks
2002-04-21 19:04:26 +00:00
Philippe Charnier
e8937ba009 Use `The .Nm utility' 2002-04-20 12:18:28 +00:00