least glib, glib2, and pkgconfig.
2004-02-26 Richard Earnshaw <rearnsha@arm.com>
Daniel Jacobowitz <drow@mvista.com>
PR target/14302
* arm.h (ARM_GO_IF_LEGITIMATE_ADDRESS): Don't check the mode
size for minipool references.
Many thanks to Richard Earnshaw for close to instant copy of the gcc patch.
Bump gcc version to "3.3.3 (NetBSD nb3 20040520)"
conversions and generally avoid 128 bit (emulated) floats in between.
This fixes heaps of problems in perl >= 5.7.1, where integer values
and unsigned values are separated by the interpreter.
Thu Feb 26 18:40:40 2004 UTC by ebotcazou:
* config/sparc/sparc-protos.h (sparc_emit_floatunsdi): Add 'mode'.
(sparc_emit_fixunsdi): New prototype.
* config/sparc/sparc.c (sparc_emit_floatunsdi): Use 'mode' argument.
(sparc_emit_fixunsdi): New function.
* config/sparc/sparc.md (floatunsdisf2): Use 'general_operand' for
operand 1. Pass SFmode to sparc_emit_floatunsdi.
(floatunsdidf2): Use 'general_operand' for operand 1. Pass DFmode
to sparc_emit_floatunsdi.
(fixuns_truncsfdi2): New expander.
(fixuns_truncdfdi2): Likewise.
2003-07-10 Kazu Hirata <kazu@cs.umass.edu>
PR c/11449
* fold-const.c (sign_bit_p): Return EXP if VAL is the sign bit
of HOST_WIDE_INT.
(fold_single_bit_test): If sign_bit_p() fails, assume that the
bit being tested is not a sign bit.
2003-07-04 Jeff Law <law@redhat.com>
PR c/11428
* expr.c (do_store_flag): Pass in the correct result type
when calling fold_single_bit_test.
* fold-const.c (fold_single_bit_test): Use result_type for the
result when folding a sign bit test.
2003-07-02 Jeff Law <law@redhat.com>
* expr.c (do_store_flag): Remove special case folding for
single bit tests. Instead call back into the commonized folder
routine.
* fold-const.c (fold_single_bit_test): New function, mostly
extracted from do_store_flag, with an additional case extracted
from fold.
(fold): Call fold_single_bit_test appropriately.
* tree.h (fold_single_bit_test): Prototype.
and exception handling have a chance of working properly.
- creates libgcc, libgcc_eh and libgcc_s
- updates LIBGCC_SPEC to use them appropriately.
There's a hack in here at the moment with respect to libgcc_so in that it
is preferable to link against libgcc_so will only when -shared-libgcc is
specified (the c++ frontend does this automatically.) Configurations where
LINK_EH_SPEC is defined already do this. The gcc configuration for
NetBSD/alpha and another NetBSD platform (I forget which) actually define
LINK_EH_SPEC probably by accident rather than design.
- updates share/mk to use the compiler's knowledge of what needs linking into
libraries and executables. This removes an hppa hack.
- updates the sets for the newly created libgcc* files.
- support for linking against the _pg version of libgcc has been removed.
- Disable symbol versioning (for now)
- Make sure that libiberty knows its being configured with a
cross compiler.
- The CXX_* variables are no longer needed/used.
- LIB2FUNCS_EXTRA gets pulled in via LIB2ADD
- Get LIB1ASMFUNCS and LIB2ASMSRC
- MAYBE_USE_COLLECT2 got renamed to USE_COLLECT2 (but might not
be used)
- Get EXTRA_HEADERS so that we get generate the right paths for
CPPFLAGS
- Get some variables related to shared libgcc
PR optimization/13037
* loop.c (update_giv_derive): Ignore redundant sets of a biv
when calculating how to derive a giv from a biv.
This fixes the underlying problem in toolchain/23002.
So the problem is thus:
* The {u,}mulsidi3 generate two parallel sets which modify the upper and lower
halves of the target register.
* life_analysis() does not track subregister modifications -- if you don't
modify the whole register with a single set, it considers the register
unused.
The simple, if klugy, solution to this is to stick an explicit clobber in. It
seems to work.
While doing this, I noticed that constant folding was not happening for
32x32->64 multiplies. This is because the parallel set generated by
{u,}mulsidi3 cannot be folded at all. To solve this, I first expand to a
normal multiply, and then use a define_insn_and_split to convert it to the
parallel set after CSE and constant folding.
This patch has also been submitted to GCC bugzilla, but who knows if I'll get
a reply to that.
respective .y file has a newer timestamp (usually due to cvs checkout issues);
the code isn't .OBJDIR friendly, and it's unnecessary in our build envrionment.
abs