Below is a list of problems identified during the GDB 5.0 release
cycle. People hope to have these problems fixed in 5.1.
---
-
-Hardware watchpint problems on x86 OSes, including Linux:
-
-1. Delete/disable hardware watchpoints should free hardware debug
-registers.
-2. Watch for different values on a viariable with one hardware debug
-register.
-
-According to Eli Zaretskii <eliz@delorie.com>:
-
-These are not GDB/ia32 issues per se: the above features are all
-implemented in the DJGPP port of GDB and work in v5.0. Every
-x86-based target should be able to lift the relevant parts of
-go32-nat.c and use them almost verbatim. You get debug register
-sharing through reference counts, and the ability to watch large
-regions (up to 16 bytes) using multiple registers. (The required
-infrastructure in high-level GDB application code, mostly in
-breakpoint.c, is also working since v5.0.)
-
---
-
-RFD: infrun.c: No bpstat_stop_status call after proceed over break?
-http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00665.html
-
-GDB misses watchpoint triggers after proceeding over a breakpoint on
-x86 targets.
-
---
-
-x86 linux GDB and SIGALRM (???)
-http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html
-
-This problem has been fixed, but a regression test still needs to be
-added to the testsuite:
-http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00309.html
-
-Mark
-
---
-
-Can't build IRIX -> arm GDB.
-http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00356.html
+-- 2001-03-08
-David Whedon writes:
-> Now I'm building for an embedded arm target. If there is a way of turning
-> remote-rdi off, I couldn't find it. It looks like it gets built by default
-> in gdb/configure.tgt(line 58) Anyway, the build dies in
-> gdb/rdi-share/unixcomm.c. SERPORT1 et. al. never get defined because we
-> aren't one of the architectures supported.
+Update GDB's coding standard documentation. Known topics:
---
+ o alloca/malloc et.al.
-Problem with weak functions
-http://sourceware.cygnus.com/ml/gdb/2000-05/msg00060.html
+ o typedef and structs
-Dan Nicolaescu writes:
-> It seems that gdb-4.95.1 does not display correctly the function when
-> stoping in weak functions.
->
-> It stops in a function that is defined as weak, not in the function
-> that is actually run...
+ o ISO-C
---
+and most likely also:
-GDB 5.0 doesn't work on Linux/SPARC
+ o include conventions
--
-Thread support. Right now, as soon as a thread finishes and exits,
-you're hosed. This problem is reported once a week or so.
+Obsolete ser-ocd.c.
--
--
-Re: GDB 5.0.1?
-http://sources.redhat.com/ml/gdb/2000-07/msg00038.html
+x86 linux GDB and SIGALRM (???)
+http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html
+
+This problem has been fixed, but a regression test still needs to be
+added to the testsuite:
+http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00309.html
+
+Mark
+
+[The test has been submitted for approval - cagney]
+
+--
-Is the Solaris 8 x86 problem fixed? When you configure it, configure
-incorrectly determines that I have no curses.h. This causes mucho
-compilation errors later on.
+RFD: infrun.c: No bpstat_stop_status call after proceed over break?
+http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00665.html
+
+GDB misses watchpoint triggers after proceeding over a breakpoint on
+x86 targets.
-Simply editing the config.h to define CURSES_H fixes the problem, and
-then the build works fine.
+--
-The status for this problem:
+GDB 5.0 doesn't work on Linux/SPARC
-Solaris 8 x86 (PIII-560)
-gcc 2.95.2
+There are two parts to this.
-I had the same problem with several of the snapshots shortly before
-5.0 became official, and 5.0 has the same problem.
+ o GDB 5.0 doesn't work on GNU/Linux/SPARC32
-I sent some mail in about it long ago, and never saw a reply.
+ o GDB 5.0 doesn't work on the new target
+ GNU/Linux/SPARC64
-I haven't had time to figure it out myself, especially since I get all
-confused trying to figure out what configure does, I was happy to find
-the workaround.
+GDB does build on both these targets.
-Mike
+The first problem is the one that should be fixed.
--
Enable MI by default. Old code can be deleted after 5.1 is out.
+Issues:
+
+ o syntax change where a list would
+ look like:
+ [ foo=a, foo=b, foo=c ]
+ instead of
+ { foo=a, foo=b, foo=c }
+
+ o kill off the idea of a reverse
+ query.
+
+ o review test cases
+
+ o enable it
+
--
Pascal (Pierre Muller, David Taylor)
Indent -gnu ?
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00496.html
+[I think this has been merged, need to confirm - cagney]
+
--
Java (Anthony Green, David Taylor)
Patch: handle N_MAIN stab
http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00527.html
---
+-- 2001-03-08
-[Comming...]
+Add CRIS target.
-Modify gdb to work correctly with Pascal.
-
---
-
-Revised UDP support (was: Re: [Fwd: [patch] UDP transport support])
-http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00000.html
-
-(Broken) support for GDB's remote protocol across UDP is to be
-included in the follow-on release.
-
-It should be noted that UDP can only work when the [Gg] packet fits in
-a single UDP packet.
-
-There is also much debate over the merit of this.
+A predicate to this is the multi-arching of SOFTWARE_SINGLE_STEP(). A
+patch has been submitted.
--
The following code cleanups will hopefully be applied to GDB 5.1.
---
+-- 2001-03-26
-Delete macro TARGET_BYTE_ORDER_SELECTABLE.
+Resolve the build status of all broken targets as identified by the
+MAINTAINERS file.
-Patches in the database.
+ o arm-* vs NetBSD's lack of ``unix''
+ o arm-* vs IRIX (see below)
--
--
-Purge PARAMS.
-
-Eliminate all uses of PARAMS in GDB's source code.
+ GDB 5.1 - Known Problems
+ ========================
--
-printcmd.c (print_address_numeric):
+z8k
-NOTE: This assumes that the significant address information is kept in
-the least significant bits of ADDR - the upper bits were either zero
-or sign extended. Should ADDRESS_TO_POINTER() or some
-ADDRESS_TO_PRINTABLE() be used to do the conversion?
+The z8k has suffered bit rot and is known to not build. The problem
+was occuring in the opcodes directory.
--
-Compiler warnings.
+m88k
-Eliminate all warnings for at least one host/target for the flags:
--Wimplicit -Wreturn-type -Wcomment -Wtrigraphs -Wformat -Wparentheses
--Wpointer-arith -Wuninitialized
+The m88k has suffered bit rot and is known to not build.
--
-Follow through `make check' with --enable-shared.
+Solaris 8 x86 CURSES_H problem
+http://sources.redhat.com/ml/gdb/2000-07/msg00038.html
-When the srcware tree is configured with --enable-shared, the `expect'
-program won't run properly. Jim Wilson found out gdb has a local hack
-to set LD_LIBRARY_PATH, but, AFAIK, no other project has been hacked
-similarly.
+The original problem was worked around with:
-http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00845.html
+ 2000-06-06 Michael Snyder <msnyder@cygnus.com>
+
+ * configure.in: Enable autoconf to find curses.h on Solaris 2.8.
+ * configure: Regenerate.
+
+When building both GDB and SID using the same source tree the problem
+will still occure. sid/component/configure.in mis-configures
+<curses.h> and leaves wrong information in the config cache.
--
--
-Fix at least one thread bug.
+Thread support. Right now, as soon as a thread finishes and exits,
+you're hosed. This problem is reported once a week or so.
--
--
-Objective C/C++ Support. Bu hopefully sooner...
+GCC 3.0 ABI support (but hopefully sooner...).
+
+--
+
+Objective C/C++ support (but hopefully sooner...).
--
--
-Eliminate more compiler warnings.
+Remove old code that does not use ui_out functions and all the related
+"ifdef"s. This also allows the elimination of -DUI_OUT from
+Makefile.in and configure.in.
+
+--
+
+Compiler warnings.
+
+Eliminate warnings for all targets on at least one host for one of the
+-W flags. Flags up for debate include: -Wswitch -Wcomment -trigraphs
+-Wtrigraphs -Wunused-function -Wunused-label -Wunused-variable
+-Wunused-value -Wchar-subscripts -Wtraditional -Wshadow -Wcast-qual
+-Wcast-align -Wwrite-strings -Wconversion -Wstrict-prototypes
+-Wmissing-prototypes -Wmissing-declarations -Wredundant-decls
+-Woverloaded-virtual -Winline
+
+--
+
+Deprecate, if not delete, the following:
+
+ register[]
+ register_valid[]
+ register_buffer()
+ REGISTER_BYTE()
+ Replaced by, on the target side
+ supply_register()
+ and on core-gdb side:
+ {read,write}_register_gen()
+ Remote.c will need to use something
+ other than REGISTER_BYTE() and
+ REGISTER_RAW_SIZE() when unpacking
+ [gG] packets.
+
+ STORE_PSEUDO_REGISTER
+ FETCH_PSEUDO_REGISTER
+ Now handed by the methods
+ gdbarch_{read,write}_register()
+ which sits between core GDB and
+ the register cache.
+
+ REGISTER_CONVERTIBLE
+ REGISTER_CONVERT_TO_RAW
+ REGISTER_CONVERT_TO_VIRTUAL
+ I think these three are redundant.
+ gdbarch_register_{read,write} can
+ do any conversion it likes.
+
+ REGISTER_VIRTUAL_SIZE
+ MAX_REGISTER_VIRTUAL_SIZE
+ REGISTER_VIRTUAL_TYPE
+ I think these can be replaced by
+ the pair:
+ FRAME_REGISTER_TYPE(frame, regnum)
+ REGISTER_TYPE(regnum)
+
+ DO_REGISTERS_INFO
+ Replace with
+ FRAME_REGISTER_INFO (frame, ...)
+
+ REGISTER_SIM_REGNO()
+ If nothing else rename this so that
+ how it relates to rawreg and the
+ regnum is clear.
+
+ REGISTER_BYTES
+ The size of the cache can be computed
+ on the fly.
+
+ IS_TRAPPED_INTERNALVAR
+ The pseudo registers should eventually make
+ this redundant.
+
+--
+
+Obsolete the targets.
+
+arm*-wince-pe
+mips*-*-pe
+sh*-*-pe
+
+Obsolete the protocols:
+
+RDB
+
+``As of version 5.3, WindRiver has removed the RDB server (RDB
+protocol support is built into gdb).'' -- Till.
--
The current convention is (kind of) to use $(<header>_h) in all
dependency lists. It isn't done in a consistent way.
+--
+
+ GDB 5.2 - Known Problems
+ ========================
+
--
Code Cleanups: General
--
+Investigate changing --target=a29k-amd-udi to a29k-*-coff* and
+rationalize *.mt files. The got-ya is in remote-eb.c - it has its own
+custom tty manipulation - it should be using the serial object.
+
+--
+
+Rename read_register{,_pid}() to read_unsigned_register{,_pid}().
+
+--
+
+Can't build IRIX -> arm GDB.
+http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00356.html
+
+David Whedon writes:
+> Now I'm building for an embedded arm target. If there is a way of turning
+> remote-rdi off, I couldn't find it. It looks like it gets built by default
+> in gdb/configure.tgt(line 58) Anyway, the build dies in
+> gdb/rdi-share/unixcomm.c. SERPORT1 et. al. never get defined because we
+> aren't one of the architectures supported.
+
+--
+
+Problem with weak functions
+http://sourceware.cygnus.com/ml/gdb/2000-05/msg00060.html
+
+Dan Nicolaescu writes:
+> It seems that gdb-4.95.1 does not display correctly the function when
+> stoping in weak functions.
+>
+> It stops in a function that is defined as weak, not in the function
+> that is actually run...
+
+--
+
+Follow through `make check' with --enable-shared.
+
+When the srcware tree is configured with --enable-shared, the `expect'
+program won't run properly. Jim Wilson found out gdb has a local hack
+to set LD_LIBRARY_PATH, but, AFAIK, no other project has been hacked
+similarly.
+
+http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00845.html
+
+--
+
+Delete macro TARGET_BYTE_ORDER_SELECTABLE.
+
+Patches in the database.
+
+--
+
+printcmd.c (print_address_numeric):
+
+NOTE: This assumes that the significant address information is kept in
+the least significant bits of ADDR - the upper bits were either zero
+or sign extended. Should ADDRESS_TO_POINTER() or some
+ADDRESS_TO_PRINTABLE() be used to do the conversion?
+
+--
+
The BFD directory requires bug-fixed AUTOMAKE et.al.
AUTOMAKE 1.4 incorrectly set the TEXINPUTS environment variable. It
--
-Replace strsave() + mstrsave() with libiberty:xstrdup().
-
---
-
Replace savestring() with something from libiberty.
An xstrldup()? but that would have different semantics.
--
-Eliminate mmalloc() from GDB.
+Eliminate mmalloc(), mstrsave() et.al. from GDB.
Also eliminate it from defs.h.
--
-Add __LINE__ and __FILE__ to internal_error().
-
---
-
GDB probably doesn't build on FreeBSD pre 2.2.x
http://sourceware.cygnus.com/ml/gdb-patches/2000-05/msg00378.html
exercised by the testsuite (as determined using some type of code
coverage analysis).
+--
+
+Replace the file gdb/CONTRIBUTE with a file that is generated from the
+gdb/doc/*.texinfo directory.
+
+--
+
+Rewrite/break up sparcl-tdep.c so that it uses ser*.c as the mechanism
+for accessing either the serial or UDP port.
+
--
New Features and Fixes
--
+Hardware watchpoint problems on x86 OSes, including Linux:
+
+1. Delete/disable hardware watchpoints should free hardware debug
+registers.
+2. Watch for different values on a viariable with one hardware debug
+register.
+
+According to Eli Zaretskii <eliz@delorie.com>:
+
+These are not GDB/ia32 issues per se: the above features are all
+implemented in the DJGPP port of GDB and work in v5.0. Every
+x86-based target should be able to lift the relevant parts of
+go32-nat.c and use them almost verbatim. You get debug register
+sharing through reference counts, and the ability to watch large
+regions (up to 16 bytes) using multiple registers. (The required
+infrastructure in high-level GDB application code, mostly in
+breakpoint.c, is also working since v5.0.)
+
+--
+
Add built-by, build-date, tm, xm, nm and anything else into gdb binary
so that you can see how the GDB was created.
--
-Get the TUI working on all platforms.
+Fix TUI
+
+ o readline/*.h bitrot
+
+ The TUI isn't up-to-date with
+ respect to the readline currently
+ bundled with GDB. Importing a
+ new readline is on the 5.1 wish
+ list so this can only get worse.
+
+ Grep for things like term_cursor_move.
+
+ (To be honest, I don't see anyone
+ importing a new readline before 5.1 is
+ out)
+
+ o tui.c:va_catch_errors() bitrot
+
+ This nasty piece of work used knowledge
+ of the internals of GDBs error functions :-(
+ Ever since those internals were cleaned
+ up this code has been broken. :-(
+
+ o tuiWin.c:c_makeVisibleWithNewHeight() broken
+ tuiLayout.c:_extractDisplayStartAddr() broken
+
+ Both these function call find_line_pc()
+ incorrectly (wrong args, wrong return value).
+
+ I suspect this bug has always been there!
+ It had been hidden because those files
+ didn't include the necessary header files
+ from gdb proper :-(
+
+ o tuiRegs() host dependant
+
+ Not suprisingly, this isn't a very portable
+ section of code. However, I'm sure people
+ could live with no regs in the short to
+ medium term.
+
+ o defs.h: #include "tui.h" et.al.
+
+ I'm not sure where this came from.
+ It was a really bad idea.
+
+ To get things to compile I did a nasty
+ hack (Just declare what was needed and
+ replace any expressions like xx->y.z()
+ in GDB proper with function calls). I
+ could commit it slightly cleaned up if
+ you like.
+
+ Medium Term. the #ifdef TUI and TuiDo()
+ should be changed to hooks (like GDBTK).
+ The gdb-events.[hc] is there for that
+ purpose (1)
+
+ o tui.c:_tuiReset() host dependant
+
+ tui.c contains a lump of termio[s]
+ I suspect an equivalent block of
+ code can be lifted from readline.
+ An equivalent readline function may
+ even be available.
+
+ o curses.h vs ncurses.h.
+
+ Simple portability problem.
+
+ o subsetCompare()
+
+ This function is a mystery - where is it?
+
+ o tui-file.[hc] cleanup
+
+ This can be significantly simplified.
+
+ o The code should be pacified. (-Werror -W...)
+
+ There are plenty of #includes,
+ duplicate #includes, missing function decls
+ and the like.
+
+ Some of the problems I found were through
+ fixing a few of the warnings.
+
+ o The code should be GNUtified.
+
+ It would be very nice to have this code
+ look like the rest of GDB. That way people
+ would be more accepting of it as a true
+ gdb component.
+
+ Until it is GNUtified it is going to stick
+ out like a sore thumb to the programmer.
+
+ o The code should be clearly copyrighted
+
+ (FSF, with due credit to HP)
--
> practical matter, the current thread support is somewhat more annoying
> than when GDB was thread-unaware.
---
-
-Migrate qfThreadInfo packet -> qThreadInfo. (Andrew Cagney)
-
-Add support for packet enable/disable commands with these thread
-packets. General cleanup.
-
-[PATCH] Document the ThreadInfo remote protocol queries
-http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00832.html
-
-[PATCH] "info threads" queries for remote.c
-http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00831.html
-
--
Language Support
--
+Revised UDP support (was: Re: [Fwd: [patch] UDP transport support])
+http://sourceware.cygnus.com/ml/gdb-patches/2000-04/msg00000.html
+
+(Broken) support for GDB's remote protocol across UDP is to be
+included in the follow-on release.
+
+It should be noted that UDP can only work when the [Gg] packet fits in
+a single UDP packet.
+
+There is also much debate over the merit of this.
+
+--
+
+Migrate qfThreadInfo packet -> qThreadInfo. (Andrew Cagney)
+
+Add support for packet enable/disable commands with these thread
+packets. General cleanup.
+
+[PATCH] Document the ThreadInfo remote protocol queries
+http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00832.html
+
+[PATCH] "info threads" queries for remote.c
+http://sourceware.cygnus.com/ml/gdb-patches/2000-q1/msg00831.html
+
+--
+
Remote protocol doco feedback.
Too much feedback to mention needs to be merged in (901660). Search
--
+Fix the ``!'' packet.
+
+JT reported that the existing targets do, in fact return ``OK'' so it
+is possible to merge remote and extended-remote targets.
+
+--
+
+Drop ``<address>'' from the [SsCc] packets.
+
+I don't think that GDB generates them so having it in the protocol is
+silly.
+
+--
+
+Fix doco on the ``q'' packet.
+
+It has evolved into a generic RPC. The notes should reflect this and,
+perhaps, the ``Q'' packet can be deprecated.
+
+The doco should mention that ``OK'' is a valid packet response.
+
+The doco should explain why ``OK'' needs to be a valid packet
+response.
+
+--
+
Add the cycle step command.
http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00237.html
Underlying problem is that the register file is target endian. If the
target endianess changes gdb doesn't know.
+Suggest encoding registers as NN!VALUE.
+
--
-Rename read_register{,_pid}() to read_unsigned_register{,_pid}().
+GDB should allow incomming packets to be larger than outgoing ones. A
+fully loaded T packet (containing all registers) can be very large -
+definitly larger than a corresponding Gg packet.
--
After that it needs to be multi-arched and made independent of any
host signal numbering.
+Once this is done, the signal enum can probably be moved to
+include/gdb so that it is available to embedded stubs.
+
--
Update ALPHA so that it uses ``struct frame_extra_info'' instead of
projects / deliverable / binutils-gdb.git / blobdiff