Thread: [Basilisk-cvs] BasiliskII/src/Unix sigsegv.cpp

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- fixed "SIGSEGV recovery" not supported case

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- use "volatile" to prevent from optimization for writes to page

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- small fixes for Linux/ia64

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- Add EIP recovery on Linux/x86 when using siginfo_t based signal handlers

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
direct addressing works under FreeBSD (says Michael Alyn Miller :-)

On Fri, 1 Feb 2002, CVS wrote:

> direct addressing works under FreeBSD (says Michael Alyn Miller :-)

Ah? Which version of FreeBSD? Last time I tried with FreeBSD 4.1, if
memory serves, the sigsegv.* tests passed when manually compiled (using
SIGBUS instead of SIGSEGV, of course). However, they failed when run
inside the configure script. By failure, I mean the program doesn't
terminate, thus blocking configure.

Bye,
Gwenole.

On Sat, 2 Feb 2002, Gwenole Beauchesne wrote:

> On Fri, 1 Feb 2002, CVS wrote:
>
> > direct addressing works under FreeBSD (says Michael Alyn Miller :-)
>
> Ah? Which version of FreeBSD? Last time I tried with FreeBSD 4.1, if
> memory serves, the sigsegv.* tests passed when manually compiled (using
> SIGBUS instead of SIGSEGV, of course). However, they failed when run
> inside the configure script. By failure, I mean the program doesn't
> terminate, thus blocking configure.

I tested the change with 4.3-RELEASE.  It should be forwards compatible.
There may have been other issues in 4.1 that caused it to fail before..?

Let me know if you have any other questions.

Michael Alyn Miller

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- Add fault instruction recovery for Linux/ppc and siginfo_t handlers

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- Merge with clisp SIGSEGV library version 1.2 with support added for
  HP-UX (contributed by Paolo Bonzini)
- Add missing? include <ucontext.h> for Irix and OSF/1 support

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Implement "ignoresegv" feature on Linux/ppc:
- Unix/sigsegv.cpp (powerpc_skip_instruction): New from SheepShaver code.
- README (ignoresegv): Add Linux/ppc to list of supported platforms.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Implement "ignoresegv" feature on Darwin/ppc (tested on MacOS X)
- Unix/sigsegv.cpp (powerpc_decode_instruction): New.
  (get_fault_instruction [MacOS X]): Factorize code.
  (get_fault_instruction [Linux/m68k]): Don't use expression statement.
- README (ignoresegv): Add Darwin/ppc to list of supported platforms.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- sigsegv.cpp (ix86_skip_instruction): Fix DEBUG compilation.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- Fix compilation on Linux/x86 with SIGCONTEXT_SUBTERFUGE

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Implement "ignoresegv" feature on FreeBSD/x86 (tested on FreeBSD 4.5)
- sigsegv.cpp (ix86_skip_instruction): Add decoder for movzwl instructions.
  (main): oddly, FreeBSD doesn't seem to let a write to a page if it is
  write-only. Aka. make the page readable too.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Backout "ignoresegv" support on FreeBSD/x86 for now. Unfortunately, the
configure script would hang whereas standalone testing will pass all tests.
Any idea why??
- Unix/sigsegv.cpp (ix86_skip_instruction): Add decoder for mozbl instruction.
  (SIGSEGV_REGISTER_FILE [FreeBSD/x86]): Note why we start at EDI offset
  rather than plain sigcontext pointer. i.e. I don't know sigset_t size
  beforehand and don't intend to.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Correctly guard FreeBSD bits

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
- New arch support for IP retrieval on SIGSEGV.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
configure script is reportedly no longer crashing on FreeBSD when
SIGSEGV_SKIP_INSTRUCTION is set.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Bring x86 instruction skipper back to life

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
fix typo

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
really fix typo!

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Mach memory fault recovery from Michael Z. Sliczniak

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
Make sure reported fault instruction address falls into the expected code
range. This is a configure-time safety net that depends on the GCC "Labels
as Values" extension.

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
return something more convenient to debug on errors

Update of /cvs/BasiliskII/src/Unix

Modified Files:
	sigsegv.cpp 
Log Message:
indentation fixes, optimize handle_badaccess() to receive only necessary
data and don't copy a struct sigcontext again on x86.

Thread: [Basilisk-cvs] BasiliskII/src/Unix sigsegv.cpp

basilisk-cvs