old-cross-binutils/gas
Jeff Law 48ac0c90e9 * config/tc-hppa.c (pa_space): Do not report an error for a .space
directive which does not define a "well-known" space and does
	not include a space number as an argument.

	* config/tc-hppa.c (pa_def_subspaces): Correct initialization of the
	"defined", "loadable", "code_only" and "space_index" fields.
	(pa_def_spaces): Correct initialization of the "spnum", "defined",
	and "private" fields.

	* write.h (struct fix): Add new tc_fix_data field for the various
	backends to attach machine dependent fixup information to.
	* write.c (fix_new_internal): Initialize new tc_fix_data field.
	* config/tc-hppa.c (hppa_fix_struct): Delete unnecessary fix_fixP and
	fx_next fields.
	(hppa_find_hppa_fix): Delete unnecessary function.  Fix all
	callers	to get HPPA fixup information from the tc_fix_data field
	in the GAS fixup.
	(hppa_fix_root): Delete unnecessary variable.
	(fix_new_hppa): Attach HPPA fixup data to the GAS fixup.

	* config/tc-hppa.c (pa_set_start_symbol); Delete unwanted
	function.  Fix all callers.
	(subspace_dictionary_chain): Delete unused ssd_start_sym field.

	* config/tc-hppa.c (hppa_fix_adjustable): New function to determine
	if a particular fixup is adjustable.
	* config/tc-hppa.h (tc_fix_adjustable): Call hppa_fix_adjustable to
	perform the real work.

	* config/tc-hppa.h (RELOC_EXPANSION_POSSIBLE): Move definition out
	of OBJ_XXX conditionals.
	(MAX_RELOC_EXPANSION): Likewise.

	* config/tc-hppa.c (log2): Renamed from is_power_of_2.  Fix all
	callers. Now returns log2 (N) for positive N which are an exact
	power of two or -1 for an error.

	* config/tc-hppa.c (pa_callinfo): Range check values provided for
	ENTRY_GR, ENTRY_FR and ENTRY_SR.  Properly adjust vaues before
	inserting them into the unwind table.

	* config/tc-hppa.c (NEEDS_FIXUP): Delete definition and all references.
	(hppa_gen_reloc_type): New object format dependent macro.
	(pa_ip): Delete tons of code which was either OBJ_SOM or OBJ_ELF
	conditional.  The code can (and will) be shared between SOM & ELF
	formats in the near future.
	(cons_fix_new_hppa, md_apply_fix_1): Likewise.
	(pa_build_unwind_subspace, process_exit, pa_exit): Likewise.
	(tc_gen_reloc): Use hppa_gen_reloc rather than an object format
	specific call.

	* config/tc-hppa.c (pa_comm): Set the segment for a common symbol
	to bfd_und_section.

	* config/obj-elf.h (obj_elf_version): Add extern prototype.

	* configure.in (hppa-*-bsd*): New configuration.
	BFD is always used for GAS generating SOM objects.

	* write.c (adjust_reloc_syms): Set sy_used_in_reloc if an
	adjustment is rejected by the target machine.

	* config/tc-hppa.c (pa_big_cons): Delete function and its
	declaration.  All callers changed to use pa_cons.

	* write.c (fixup_segment): Fix indention and open/close brace
	problem.

	From Pete Hoogenboom:
	* config/tc-hppa.c (md_atof): Return a NULL on success rather than
	an empty string.
1993-10-28 20:49:19 +00:00
..
config Merge with current Utah code: 1993-10-28 20:43:13 +00:00
doc list a bunch of files known to get deleted 1993-10-26 17:30:47 +00:00
testsuite bunch of changes... 1993-10-25 16:48:51 +00:00
.gdbinit stop in abort 1993-02-18 02:55:47 +00:00
.Sanitize clean v9 stuff out of configure.in 1993-06-04 23:06:28 +00:00
app.c gcc lint. See ChangeLog for details. Also: 1993-09-10 16:01:07 +00:00
as.c localize a conditionally used variable 1993-07-07 12:33:11 +00:00
as.h * subsegs.c: Renamed non-BFD_ASSEMBLER subseg_new to subseg_set. 1993-09-13 21:32:07 +00:00
atof-generic.c White space and comment changes. #ifdef __STDC__ becomes #if __STDC__ 1992-02-15 21:13:03 +00:00
bignum-copy.c More gcc lint with harsher warning options. 1993-09-11 00:15:39 +00:00
bignum.h Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
bit_fix.h fighting bitrot in a major way 1992-02-17 15:54:49 +00:00
ChangeLog * config/tc-hppa.c (pa_space): Do not report an error for a .space 1993-10-28 20:49:19 +00:00
ChangeLog.v9 Fix up membar argument handling. 1993-10-25 18:27:24 +00:00
cond.c White space and comments only. The devo tree prior to this delta is 1992-02-13 08:33:54 +00:00
config-gas.com Lots of changes from Eric Youngdale, mostly for VAX and VMS support. 1993-01-07 10:48:49 +00:00
configure.bat More renames from DJ the DOS martyr 1992-02-22 00:23:03 +00:00
configure.in Set bfd_gas for all sparc targets. Added facility for keeping or rejecting 1993-10-21 19:50:06 +00:00
CONTRIBUTORS record some of my gas work 1993-03-01 19:31:57 +00:00
COPYING This commit was manufactured by cvs2svn to create branch 'binutils'. 1991-04-04 18:19:54 +00:00
debug.c force all files to end in "/* end of filename" 1992-02-13 09:13:54 +00:00
expr.c gcc lint. See ChangeLog for details. Also: 1993-09-10 16:01:07 +00:00
expr.h * Changes to keep a full expression as the value of a symbol, not 1993-07-14 19:35:45 +00:00
flonum-copy.c force all files to end in "/* end of filename" 1992-02-13 09:13:54 +00:00
flonum-konst.c Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
flonum-mult.c Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
flonum.h Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
frags.c force all files to end in "/* end of filename" 1992-02-13 09:13:54 +00:00
frags.h Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
hash.c force all files to end in "/* end of filename" 1992-02-13 09:13:54 +00:00
hash.h force all files to end in "/* end of filename" 1992-02-13 09:13:54 +00:00
hex-value.c Lots of changes from Eric Youngdale, mostly for VAX and VMS support. 1993-01-07 10:48:49 +00:00
input-file.c Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
input-file.h Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
input-scrub.c Thu Sep 10 09:23:15 1992 Ian Lance Taylor (ian@cygnus.com) 1992-09-10 16:25:31 +00:00
link.cmd new gas main line 1991-04-04 18:19:55 +00:00
listing.c gcc lint. See ChangeLog for details. Also: 1993-09-10 16:01:07 +00:00
listing.h Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
make-gas.com SUB_SEGMENT_ALIGN now takes an argument. 1993-03-03 19:46:12 +00:00
Makefile.in (INCLUDES): Look in ../bfd, not ../include, for bfd.h. 1993-06-20 22:25:46 +00:00
messages.c * messages.c (as_fatal): Do mention that it's the assembler that 1993-10-11 20:17:28 +00:00
NOTES updated to-do list 1993-08-18 20:45:18 +00:00
NOTES.config new gas main line 1991-04-04 18:19:56 +00:00
obj.h Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
output-file.c force all files to end in "/* end of filename" 1992-02-13 09:13:54 +00:00
output-file.h Ran "indent", for GNU coding style; some code & comments still need fixup. 1992-11-23 20:39:57 +00:00
read.c Changes to let cons handle bignums like general expressions. 1993-10-06 17:31:31 +00:00
read.h gcc lint. See ChangeLog for details. Also: 1993-09-10 16:01:07 +00:00
README This commit was manufactured by cvs2svn to create branch 'binutils'. 1991-04-04 18:19:54 +00:00
README-quirks no longer using gas-format.el 1992-12-03 23:15:59 +00:00
README-vms vms stuff from fsf version - untested 1992-12-03 23:38:49 +00:00
README.coff new gas main line 1991-04-04 18:19:56 +00:00
README.rich new gas main line 1991-04-04 18:19:56 +00:00
struc-symbol.h * Changes to keep a full expression as the value of a symbol, not 1993-07-14 19:35:45 +00:00
subsegs.c * subsegs.c: Renamed non-BFD_ASSEMBLER subseg_new to subseg_set. 1993-09-13 21:32:07 +00:00
subsegs.h force all files to end in "/* end of filename" 1992-02-13 09:13:54 +00:00
symbols.c * subsegs.c: Renamed non-BFD_ASSEMBLER subseg_new to subseg_set. 1993-09-13 21:32:07 +00:00
symbols.h * Changes to keep a full expression as the value of a symbol, not 1993-07-14 19:35:45 +00:00
tc.h BFD_ASSEMBLER md_convert_frag section arg is not a pointer 1993-05-05 20:46:11 +00:00
version.c bump main trunk to version 2.0.1 1993-03-09 15:44:37 +00:00
write.c * write.c (write_relocs): For relocs that are pc_relative and 1993-10-25 17:49:24 +00:00
write.h * subsegs.c: Renamed non-BFD_ASSEMBLER subseg_new to subseg_set. 1993-09-13 21:32:07 +00:00
xmalloc.c More gcc lint with harsher warning options. 1993-09-11 00:15:39 +00:00

This is the beta-test version of the GNU assembler.  (Probably
around Version 1.38, but check version.c which gets updated more
often than this readme.)

These files are currently set up to allow you to compile all of the
versions of the assembler on the same machine.  'make all' compiles
all of them.  The resulting executable names are:

	68020		a68
	Vax		avax
	NS 32xxx	a32k
	Intel 80386	a386
	SPARC		asparc
	AMD 29000	asm29k

The Makefile contains instructions on how to make one of the
assemblers compile as the default.

Before you can compile the 68020 version of the assembler, you must
make m68k.h be a link to m-sun3.h , m-hpux.h or m-generic.h .  If
you are on a SUN-3 (or other machine that uses a magic number of
(2 << 16) | OMAGIC type 'ln -s m-sun3.h m68k.h' else if you are on a
machine running HP-UX, type 'ln m-hpux.h m689k.h' else type
'ln -s m-generic.h m68k.h' If your machine does not support symbolic
links, omit the '-s'.

See the instructions in the Makefile for compiling gas for the Sequent
Symmetry (dynix 3.0.12 + others?) or for the HP 9000/300

If your machine does not have both varargs.h and vfprintf(), but does have
_doprnt() add -DNO_VARARGS to the CFLAGS line in the makefile.  If your
machine has neither vfprintf() or _doprnt(), you will have to change
messages.c in order to get readable error messages from the assembler.

The assembler has been modified to support a feature that is
potentially useful when assembling compiler output, but which may
confuse assembly language programmers.  If assembler encounters a
.word pseudo-op of the form symbol1-symbol2 (the difference of two
symbols), and the difference of those two symbols will not fit in 16
bits, the assembler will create a branch around a long jump to
symbol1, and insert this into the output directly before the next
label:  The .word will (instead of containing garbage, or giving an
error message) contain (the address of the long jump)-symbol2.  This
allows the assembler to assemble jump tables that jump to locations
very far away into code that works properly.  If the next label is
more than 32K away from the .word, you lose (silently); RMS claims
this will never happen.  If the -k option is given, you will get a
warning message when this happens.


	REPORTING BUGS IN GAS

Bugs in gas should be reported to bug-gnu-utils@prep.ai.mit.edu  If you can't
get through to prep, try hack@gnu.ai.mit.edu or hack@media-lab.media.mit.edu

If you report a bug in GAS, please remember to include:

A description of exactly what went wrong.

The type of machine GAS was running on (VAX, 68020, etc),

The Operating System GAS was running under.

The options given to GAS.

The actual input file that caused the problem.

It is silly to report a bug in GAS without including an input file for
GAS.  Don't ask us to generate the file just because you made it from
files you think we have access to.

1. You might be mistaken.
2. It might take us a lot of time to install things to regenerate that file.
3. We might get a different file from the one you got, and might not see any
bug.

To save us these delays and uncertainties, always send the input file
for the program that failed.

If the input file is very large, and you are on the internet, you may
want to make it avaliable for anonymous FTP instead of mailing it.  If you
do, include instructions for FTP'ing it in your bug report.

------------------------------ README.APOLLO ---------------------------------

The changes required to get the GNU C compiler running on
Apollo 68K platforms are available via anonymous ftp from
labrea.stanford.edu (36.8.0.47) in the form of a compressed
tar file named "/pub/gnu/apollo-gcc-1.37.tar.Z".
The size of the file is 84145 bytes.

To build GCC for the Apollo you'll need the virgin FSF
distributions of bison-1.03, gas-1.34, and gcc-1.37. They
are also on labrea.stanford.edu as well as prep.ai.mit.edu.
My changes are to enable gas to produce Apollo COFF object
files and allow gcc to parse some of the syntax extensions
which appear in Apollo C header files. Note that the
COFF encapsulation technique cannot be used on the Apollo.
                                                             
The tar file should be unpacked in the directory containing
the gas-1.34 and gcc-1.37 directories; a few files will be overlaid,
and an APOLLO-GCC-README file will appear in the top directory.
This file contains detailed instructions on how to proceed.

These changes will only work for SR10.1 or later systems, using
the 6.6 or later version of the Apollo C compiler.

If you do not have ftp access, I can mail you the changes in the
form of diffs; they are approximately 40K in length. If you request
them, be sure to give me a voice phone number so I can contact you
in case I can't send you mail; I've had several requests in the
past from people I can't contact.

By the way, I'm working on getting the GNU C++ compiler running;
there are a couple problems to solve. I hope to be able to announce
the Apollo version shortly after the 1.37 version is released.

John Vasta                Hewlett-Packard Apollo Systems Division
vasta@apollo.hp.com       M.S. CHA-01-LT
(508) 256-6600 x6362      300 Apollo Drive, Chelmsford, MA 01824
UUCP: {decwrl!decvax, mit-eddie, attunix}!apollo!vasta

------------------------------------

You might refer others who are interested in a similar thing.

Kevin Buchs    buchs@mayo.edu


------------------------------ README.COFF -----------------------------------

If you have a COFF system, you may wish to aquire

	UUCP: osu-cis!~/gnu/coff/gnu-coff.tar.Z
	or
	FTP:  tut.cis.ohio-state.edu:/pub/gnu/coff/gnu-coff.tar.Z

These contain patches for gas that will make it produce COFF output.
I have never seen these patches, so I don't know how well they work.