espressif/binutils-gdb
1
0
Fork 0
mirror of https://github.com/espressif/binutils-gdb.git synced 2025-06-24 04:00:07 +08:00
Code Issues Packages Projects Releases Wiki Activity

* config/m68k/tm-m68k.h (SIG_PC_FP_OFFSET, SIG_SP_FP_OFFSET):

Browse Source 
Define for correct handling of bachtraces through _sigtramp.
This commit is contained in:
Peter Schauer
1993-08-16 10:13:22 +00:00
parent adce222501
commit c67d08e89b
1 changed files with 19 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
gdb/config/m68k/tm-sun3.h
View File

@ -71,3 +71,22 @@ extern CORE_ADDR m68k_saved_pc_after_call PARAMS ((struct frame_info *));
/* The code which tries to deal with this bug is never harmful on a sun3. */ /* The code which tries to deal with this bug is never harmful on a sun3. */
#define SUN_FIXED_LBRAC_BUG (0) #define SUN_FIXED_LBRAC_BUG (0)
/* On the sun3 the kernel pushes a sigcontext on the user stack and then
`calls' _sigtramp in user code. _sigtramp saves the floating point status
on the stack and calls the signal handler function. The stack does not
contain enough information to allow a normal backtrace, but sigcontext
contains the saved user pc/sp. FRAME_CHAIN and friends in tm-m68k.h and
m68k_find_saved_regs deal with this situation by manufacturing a fake frame
for _sigtramp.
SIG_PC_FP_OFFSET is the offset from the signal handler frame to the
saved pc in sigcontext.
SIG_SP_FP_OFFSET is the offset from the signal handler frame to the end
of sigcontext which is identical to the saved sp at SIG_PC_FP_OFFSET - 4.
Please note that it is impossible to correctly backtrace from a breakpoint
in _sigtramp as _sigtramp modifies the stack pointer a few times. */
#undef SIG_PC_FP_OFFSET
#define SIG_PC_FP_OFFSET 324
#define SIG_SP_FP_OFFSET 332