From mboxrd@z Thu Jan 1 00:00:00 1970 From: John Wehle To: gdb-patches@cygnus.com Cc: shebs@cygnus.com Subject: Re: GDB 4.17 Patch for stack aligned i386 code Date: Wed, 24 Feb 1999 19:41:00 -0000 Message-ID: <199902250335.WAA27443@jwlab.FEITH.COM> X-SW-Source: 1999-02/msg00014.html Message-ID: <19990224194100.L-x5hnO_rJ4GXQeanyp9mWoc5r9ANj3mUq1wAwZijHM@z> Just curious what the status is of this patch. Anything I need to do at this point? Wed Jan 20 22:14:03 EST 1999 John Wehle (john@feith.com) * tm-i386.h (FRAME_CHAIN): Call i386_frame_chain for signal handler callers. (EXTRA_FRAME_INFO): Define stack_mask. (FRAME_ARGS_ADDRESS, FRAME_LOCALS_ADDRESS): Use it. * i386-tdep.c (i386_get_frame_setup): Set it. (i386_frame_find_saved_regs): Use it. (i386_init_extra_frame_info): Initialize it. (i386_frame_find_saved_regs): Don't accidently zero fi->fsr.regs[FP_REGNUM]. (i386_sigcontext_addr): New function. (i386v4_sigtramp_saved_pc): Use it. (i386_analyze_prologue): Use it. (i386_analyze_prologue): Handle dummy frames. (i386_analyze_prologue): If the start of a function can't be located than assume the frame is in the frame pointer. (i386_frame_chain): If the caller's prologue can't be analyze then return the current frame as the frame. (i386_frame_chain): Check the current frame for FP_REGNUM. (i386_frame_chain): Handle chaining back from a signal handler caller to a function which uses the stack pointer for the frame pointer. (i386_frame_chain): Include the stack space occupied by FP_REGNUM when caculating the caller's frame. (i386_pop_frame): Don't bother explicitly restoring PC_REGNUM, it is restored by the for loop. * tm-i386bsd.h (SIGCONTEXT_FP_OFFSET, SIGCONTEXT_SP_OFFSET): Define. * tm-i386sol2.h: Likewise. * tm-i386v4.h: Likewise. -- John ------------------------------------------------------------------------- | Feith Systems | Voice: 1-215-646-8000 | Email: john@feith.com | | John Wehle | Fax: 1-215-540-5495 | | -------------------------------------------------------------------------