From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mark Kettenis To: eliz@is.elta.co.il Cc: gdb-patches@sourceware.cygnus.com Subject: Re: [PATCH] Provide `long double' support for most i386 targets Date: Mon, 03 Apr 2000 13:05:00 -0000 Message-id: <200004032004.e33K4Lq06070@delius.kettenis.local> References: <200003262226.e2QMQQu07762@delius.kettenis.local> <200004030845.EAA11820@indy.delorie.com> X-SW-Source: 2000-04/msg00042.html Date: Mon, 3 Apr 2000 04:45:36 -0400 (EDT) From: Eli Zaretskii I really like it if code is being removed :-). > The LD_I387 definition can probably also be removed. Its > only purpose would be the HEX_LONG_DOUBLE_INPUT macro, but that macro > is never used in the GDB sources. Hmm, this HEX_LONG_DOUBLE_INPUT seems like a useful feature, but it doesn't work, including in GDB 4.18. Did it ever work? I removed the macro. I think such a feature is in HJ's version Linux-only version of gdb 4.17. The implementation there is a bit kludgey, and AFAIK never submitted as an isolated feature. Moreover I don't see why this should be i386 and/or Linux only. Probably a wish-list item. Mark >From jtc@redback.com Mon Apr 03 14:35:00 2000 From: jtc@redback.com (J.T. Conklin) To: gdb-patches@sourceware.cygnus.com Subject: FYI: patched i386/xm-nbsd.h Date: Mon, 03 Apr 2000 14:35:00 -0000 Message-id: <5mem8msw9v.fsf@jtc.redback.com> X-SW-Source: 2000-04/msg00043.html Content-length: 693 As suggested by Mark Kettenis, I've checked in the enclosed patch --jtc 2000-04-03 J.T. Conklin * config/i386/xm-nbsd.h (HOST_LONG_DOUBLE_FORMAT): Define. Index: config/i386/xm-nbsd.h =================================================================== RCS file: /cvs/src/src/gdb/config/i386/xm-nbsd.h,v retrieving revision 1.1.1.2 diff -c -3 -p -r1.1.1.2 xm-nbsd.h *** xm-nbsd.h 1999/07/07 20:13:42 1.1.1.2 --- xm-nbsd.h 2000/04/03 21:30:47 *************** *** 20,22 **** --- 20,24 ---- /* Get generic NetBSD host definitions. */ #include "xm-nbsd.h" + + #define HOST_LONG_DOUBLE_FORMAT &floatformat_i387_ext -- J.T. Conklin RedBack Networks >From jtc@redback.com Mon Apr 03 14:50:00 2000 From: jtc@redback.com (J.T. Conklin) To: Andrew Cagney Cc: Jimmy Guo , gdb-patches@sourceware.cygnus.com Subject: Re: regression analysis: gdb.sum postprocessing script Date: Mon, 03 Apr 2000 14:50:00 -0000 Message-id: <5maejasvl6.fsf@jtc.redback.com> References: <38DEFEB1.EB573B72@cygnus.com> X-SW-Source: 2000-04/msg00044.html Content-length: 1319 >>>>> "Andrew" == Andrew Cagney writes: Jimmy> Here is a Perl script I wrote to postprocess one or more Jimmy> gdb.sum files and provide a complete dejagnu test outcome Jimmy> summary (or diff if more than one gdb.sum files are given). Jimmy> It's most useful if you want to have a complete view of Jimmy> regression status between two source trees and/or if you set up Jimmy> your testing to run multiple passes (to see how a test point Jimmy> behaves in different passes), however you can use it just on Jimmy> one test run's gdb.sum file and play around with the options to Jimmy> get test outcome presented in a much organized and useful way Jimmy> than going through the flat gdb.sum file. Andrew> I have a personal preference for keeping perl scripts out of Andrew> the main GDB sources :-( Now if someone rewrote it to use Andrew> either TCL, DEJAGNU or SH then I'd have to think about it Andrew> again - they are required to run the testsuite. Jimmy, Is your script specific to the GDB testsuite or would it work with any dejagnu suite? I skimmed through the sources, and didn't see anything which obviously ties it to the GDB suite. If the script is generic, one could argue that it belongs in the deja- gnu distribution itself. --jtc -- J.T. Conklin RedBack Networks >From kevinb@cygnus.com Mon Apr 03 15:13:00 2000 From: Kevin Buettner To: gdb-patches@sourceware.cygnus.com Subject: [PATCH] Mention GNU/Linux PowerPC port in NEWS Date: Mon, 03 Apr 2000 15:13:00 -0000 Message-id: <1000403221318.ZM28000@ocotillo.lan> X-SW-Source: 2000-04/msg00045.html Content-length: 437 I just committed the following change: * NEWS (powerpc-*-linux*): Mention. Index: NEWS =================================================================== RCS file: /cvs/src/src/gdb/NEWS,v retrieving revision 1.2 diff -u -r1.2 NEWS --- NEWS 2000/04/03 13:24:08 1.2 +++ NEWS 2000/04/03 22:07:29 @@ -15,6 +15,7 @@ * New native configurations ARM GNU/Linux arm*-*-linux* +PowerPC GNU/Linux powerpc-*-linux* * New targets >From ac131313@cygnus.com Mon Apr 03 17:30:00 2000 From: Andrew Cagney To: egcs@cygnus.com Cc: gdb-patches@sourceware.cygnus.com Subject: Re: FYI: patched i386/xm-nbsd.h Date: Mon, 03 Apr 2000 17:30:00 -0000 Message-id: <38E93731.91F14979@cygnus.com> References: <5mem8msw9v.fsf@jtc.redback.com> X-SW-Source: 2000-04/msg00046.html Content-length: 918 "J.T. Conklin" wrote: > > As suggested by Mark Kettenis, I've checked in the enclosed patch > > --jtc > > 2000-04-03 J.T. Conklin > > * config/i386/xm-nbsd.h (HOST_LONG_DOUBLE_FORMAT): Define. > > Index: config/i386/xm-nbsd.h > =================================================================== > RCS file: /cvs/src/src/gdb/config/i386/xm-nbsd.h,v > retrieving revision 1.1.1.2 > diff -c -3 -p -r1.1.1.2 xm-nbsd.h > *** xm-nbsd.h 1999/07/07 20:13:42 1.1.1.2 > --- xm-nbsd.h 2000/04/03 21:30:47 > *************** > *** 20,22 **** > --- 20,24 ---- > > /* Get generic NetBSD host definitions. */ > #include "xm-nbsd.h" > + > + #define HOST_LONG_DOUBLE_FORMAT &floatformat_i387_ext Just an asside on this one. If memory serves correct, I think the long long term plan was to eliminating xm-*.h and instead use autoconf for these host side things. enjoy, Andrew >From blizzard@mozilla.org Mon Apr 03 18:28:00 2000 From: Christopher Blizzard To: gdb-patches@sourceware.cygnus.com Subject: source listing change Date: Mon, 03 Apr 2000 18:28:00 -0000 Message-id: <38E94593.668AAF44@mozilla.org> X-SW-Source: 2000-04/msg00047.html Content-length: 6459 This was forwarded to me from James Nance. He asked me to forward it on to the gdb-patches list. --Chris -------- Original Message -------- Subject: gdb patches Date: Tue, 28 Mar 2000 06:26:02 -0500 From: James Lewis Nance To: blizzard@mozilla.org Chris, As long as you are putting together patches for gdb, here is one I have been trying to get included for a while. I got it approved by Cygnus about 2 years ago, but then the person I was working up there quit. I have not been able to find anyone really interested in it since then. This patch adds a command like the "w" command in the Digital Unix ladebug debugger (though the Cygnus people had me call it "list -l"). What it does is when you type "list -l" it displays your souce code but it puts a '> ' character in front of the current line so you can find out where you are. Its one of those things you dont appreciate until you have used it a few times. Then you cant live with out it. I am attaching my patch. Please give it a try. Thanks, Jim PS: If you alias list -l to w in your .gdbinit, its much more convienient to use. diff -ru gdb-4.18.orig/gdb/ChangeLog gdb-4.18.new/gdb/ChangeLog --- gdb-4.18.orig/gdb/ChangeLog Wed Apr 7 17:34:27 1999 +++ gdb-4.18.new/gdb/ChangeLog Wed Jan 12 17:47:10 2000 @@ -1,3 +1,9 @@ +2000-01-12 Jim Nance + + * source.c symtab.h: Added a -w option to the list command. Allows + the currently executing line of code in the current stack frame to + be annotated. + 1999-04-07 Jim Blandy * GDB 4.18 released. Only in gdb-4.18.new/gdb: ChangeLog.orig Only in gdb-4.18.new/gdb: linuxthreads.c.threads diff -ru gdb-4.18.orig/gdb/source.c gdb-4.18.new/gdb/source.c --- gdb-4.18.orig/gdb/source.c Thu Dec 31 16:58:08 1998 +++ gdb-4.18.new/gdb/source.c Wed Jan 12 17:45:48 2000 @@ -976,11 +976,11 @@ /* Print source lines from the file of symtab S, starting with line number LINE and stopping before line number STOPLINE. */ -static void -print_source_lines_base (s, line, stopline, noerror) +void +print_source_lines_base (s, line, stopline, noerror, markline) struct symtab *s; int line, stopline; - int noerror; + int noerror, markline; { register int c; register int desc; @@ -1046,8 +1046,11 @@ { c = fgetc (stream); if (c == EOF) break; - last_line_listed = current_source_line; - printf_filtered ("%d\t", current_source_line++); + last_line_listed = current_source_line++; + if (last_line_listed != markline) + printf_filtered ("%d\t", last_line_listed); + else + printf_filtered ("%d >\t", last_line_listed); do { if (c < 040 && c != '\t' && c != '\n' && c != '\r') @@ -1081,7 +1084,7 @@ #if defined(TUI) if (!tui_version || m_winPtrIsNull(srcWin) || !srcWin->generic.isVisible ) - print_source_lines_base(s, line, stopline, noerror); + print_source_lines_base(s, line, stopline, noerror, 0); else { TuiGenWinInfoPtr locator = locatorWinInfoPtr(); @@ -1101,7 +1104,7 @@ tuiDo((TuiOpaqueFuncPtr)tui_vUpdateLocatorFilename, s->filename); } #else - print_source_lines_base(s, line, stopline, noerror); + print_source_lines_base(s, line, stopline, noerror, 0); #endif } @@ -1123,6 +1126,18 @@ sals->sals[i].symtab->filename, sals->sals[i].line); } +static void printw (s, line, stopline, noerror) + struct symtab *s; + int line; + int stopline; + int noerror; +{ + int nlines = lines_to_list / 2; + int bline = line > nlines ? line - nlines : 1; + int eline = line + nlines; + print_source_lines_base (s, bline, eline, noerror, line); +} + static void list_command (arg, from_tty) char *arg; @@ -1157,8 +1172,25 @@ return; } - /* "l -" lists previous ten lines, the ones before the ten just listed. */ - if (STREQ (arg, "-")) + /* We check for 2 cases here: + * "l -" lists previous ten lines, the ones before the ten just listed. + * "l -w" gives a context listing showing the current position. + */ + if (STREQN (arg, "-", 1)) + if (STREQ (arg, "-w")) + { + extern void (*print_frame_info_listing_hook) + PARAMS ((struct symtab *s, int line, int stopline, int noerror)); + void (*savefn) + PARAMS ((struct symtab *s, int line, int stopline, int noerror)); + + savefn = print_frame_info_listing_hook; + print_frame_info_listing_hook = printw; + frame_command (NULL, 0); + print_frame_info_listing_hook = savefn; + return; + } + else { if (current_source_symtab == 0) error ("No default source file yet. Do \"help list\"."); @@ -1500,7 +1532,7 @@ /* Match! */ fclose (stream); if (tui_version) - print_source_lines_base (current_source_symtab, line, line+1, 0); + print_source_lines_base (current_source_symtab, line, line+1, 0, 0); print_source_lines (current_source_symtab, line, line+1, 0); set_internalvar (lookup_internalvar ("_"), value_from_longest (builtin_type_int, @@ -1598,7 +1630,7 @@ /* Match! */ fclose (stream); if (tui_version) - print_source_lines_base (current_source_symtab, line, line+1, 0); + print_source_lines_base (current_source_symtab, line, line+1, 0, 0); print_source_lines (current_source_symtab, line, line+1, 0); set_internalvar (lookup_internalvar ("_"), value_from_longest (builtin_type_int, @@ -1695,6 +1727,7 @@ add_com ("list", class_files, list_command, concat ("List specified function or line.\n\ With no argument, lists ten more lines after or around previous listing.\n\ +\"list -w\" lists the currently executing line of souce code with context.\n\ \"list -\" lists the ten lines before a previous ten-line listing.\n\ One argument specifies a line, and ten lines are listed around that line.\n\ Two arguments with comma between specify starting and ending lines to list.\n\ diff -ru gdb-4.18.orig/gdb/symtab.h gdb-4.18.new/gdb/symtab.h --- gdb-4.18.orig/gdb/symtab.h Wed Feb 10 19:24:40 1999 +++ gdb-4.18.new/gdb/symtab.h Wed Jan 12 17:42:58 2000 @@ -1403,6 +1403,9 @@ identify_source_line PARAMS ((struct symtab *, int, int, CORE_ADDR)); extern void +print_source_lines_base PARAMS ((struct symtab *, int, int, int, int)); + +extern void print_source_lines PARAMS ((struct symtab *, int, int, int)); extern void