e than one of
functions, variables and types. (The syntax documentation is cleaner, and
I think accurate for C, but it does include some speculative future
possibilities that seemed nice when I first wrote it but I no longer think
would be a good idea.)
--
Joseph S. Myers
[EMAIL PROTECTED]
end.texi but extend.texi doesn't appear to
be mentioned in the log message). The old text only covered some of the
problem cases, but the statement that attributes were ignored in such
cases was still removed.
--
Joseph S. Myers
[EMAIL PROTECTED]
e changes in installation directories and
the addition of new configure options and directory variables, this
requires careful planning and coordination between GCC and src. (This is
definitely a matter for 4.3, not 4.2.)
--
Joseph S. Myers
[EMAIL PROTECTED]
On Fri, 20 Oct 2006, Mark Mitchell wrote:
> 2. I have not regenerated {gcc,cpplib}.pot, or sent them off to the
> translation project. Joseph, would you please do that, at your convenience?
Regeneration done. I'll submit the next 4.2 snapshot to the TP.
--
Joseph S. Myers
[EMAIL PROTECTED]
All the bugs with "4.2" in their summaries ("[4.1/4.2 Regression]" etc.)
need to have it changed to "4.2/4.3". I don't know the procedure for
this, but perhaps it needs adding to the branching checklist.
--
Joseph S. Myers
[EMAIL PROTECTED]
hould apply fine without --dry-run.
--
Joseph S. Myers
[EMAIL PROTECTED]
n audited against standard requirements.
* All uses of identifiers in the compiler audited for impact of extended
identifiers and extended identifiers implementation audited against list
in bug 9449 comment 21.
--
Joseph S. Myers
[EMAIL PROTECTED]
On Wed, 1 Nov 2006, Daniel Jacobowitz wrote:
> On Wed, Nov 01, 2006 at 06:59:42PM +0000, Joseph S. Myers wrote:
> > I'd rather simply fix glibc to work with C99 inline semantics. For the
> > above you might use
> >
> > #define tolower(c) __tolower_inline(c)
&
t, a lot of
code will simply not be fixed until -std=gnu99 becomes the default (and
accordingly that change will break a lot of code, and so should be done in
a future Stage 1).
--
Joseph S. Myers
[EMAIL PROTECTED]
foo() { ... }
exports foo, as does
inline int foo() { ... }
// Declaration below might be a previously redundant one in user code.
extern int foo();
and this is what means it's not just a straight substitution in headers.
--
Joseph S. Myers
[EMAIL PROTECTED]
vity, such as
incomplete array types (int[10] and int[5] both compatible with int[] but
not with each other), unprototyped functions and enums (two enums
compatible with the same integer type are not compatible with each other).
--
Joseph S. Myers
[EMAIL PROTECTED]
ined at a different level more like
that used by the present datastructures).
Once the datastructures are suitably general, then interfaces to them can
be considered.
--
Joseph S. Myers
[EMAIL PROTECTED]
han code. For verifying there are no
unintended changes in the compiler's behavior, compare the exact
diagnostic texts in gcc.log from test runs before and after each change.
--
Joseph S. Myers
[EMAIL PROTECTED]
what point the
local configure options --with-datarootdir --with-docdir --with-htmldir
get removed in each directory in favour of the standard options added in
2.60). 2.60 also needs automake 1.10 (and vice versa).
--
Joseph S. Myers
[EMAIL PROTECTED]
s of gcc
and src have been moved.)
--
Joseph S. Myers
[EMAIL PROTECTED]
ow (raising to a fraction with even denominator). (The
optimization from pow to cbrt is valid if you don't care about no longer
getting a NaN from a negative argument. Converting the other way (cbrt to
pow) is only OK if you don't care about negative arguments to cbrt at
all.)
--
Joseph S. Myers
[EMAIL PROTECTED]
(a*2)/2 on the basis of overflow being
undefined for a very long time, not just loops.
--
Joseph S. Myers
[EMAIL PROTECTED]
e assume wraparound signed
> integer arithmetic a la Java and LIA-1. What else can we do,
> realistically?
Convert to unsigned and do the overflow tests using unsigned arithmetic.
Encapsulate reliable overflow checks for the various arithmetic operations
and types in functions or macros in gnulib (for long long multiply, in
this case).
--
Joseph S. Myers
[EMAIL PROTECTED]
hat this optimization
was fixed to apply only to signed types and not unsigned types as well).
But if you wrap, (INT_MIN*2)/2 would be 0 not INT_MIN, for example.
--
Joseph S. Myers
[EMAIL PROTECTED]
ation is also valid for the initial operation being an
"overflow wraps" one, the optimization can be done if -fwrapv.
When individual operations can be so marked, the optimizations in question
can then be done if the original operation is either "overflow undefined"
or "overflow wraps".
--
Joseph S. Myers
[EMAIL PROTECTED]
ing-point conversions
don't have the same scope as integer arithmetic for useful optimization
based on undefined behavior.)
--
Joseph S. Myers
[EMAIL PROTECTED]
t; and non-front-end code can't use
flag_isoc99 since it isn't available in non-C-family front ends.
--
Joseph S. Myers
[EMAIL PROTECTED]
than using -fno-strict-aliasing globally: you can use unions or the
may_alias attribute.)
--
Joseph S. Myers
[EMAIL PROTECTED]
eople should convert to the unsigned variants, do computations there,
> and convert back to the signed variants. We have just promised an
> invariant that we will hold.
The suggestion is for *explicit* conversions (casts), the warnings (should
be) for implicit conversions.
--
Joseph S. Myers
[EMAIL PROTECTED]
5.4 are
equally unimpressed by the approach of recent flex releases to correctness
and compatibility.
--
Joseph S. Myers
[EMAIL PROTECTED]
s clear enough that indeterminate values may be treated
distinctly from determinate values including randomly changing at any
time.
http://www.open-std.org/jtc1/sc22/wg14/www/docs/dr_260.htm
--
Joseph S. Myers
[EMAIL PROTECTED]
applied to 4.1 and 4.2 branches. (For trunk, the
issue is fixed properly, by toplevel libgcc removing the need for
findstring in conjunction with E500 long double making -mlong-double-128 a
supported option in this case so removing the underlying ICE.)
--
Joseph S. Myers
[EMAIL PROTECTED]
d ":" as a column separator. The gcc.1
> rendering doesn't look as bad as you might think.
>
> Tested by examining the gcc.1 output. OK to install?
OK.
--
Joseph S. Myers
[EMAIL PROTECTED]
purpose of
providing deprecation warnings for incompatible changes in 4.3 (for
example, the proposed diagnostics in 4.2 for extern inline in c99/gnu99
mode); dropping a release series would require associated reversions in
mainline and delays to changes needing deprecation periods.
--
branch as well that are
small and self-contained enough that they shouldn't need to be merged
during Stage 1.)
--
Joseph S. Myers
[EMAIL PROTECTED]
wed by a set of relevant maintainers rather than using GWP".
You don't want gcc-patches clogged up with people making these indications
that someone else should review the patch but it might be of use to have
an off-list way of tracking such indications.
--
Joseph S. Myers
[EMAIL PROTECTED]
wing
<http://gcc.gnu.org/ml/gcc-patches/2007-03/msg00097.html> with the current
versions of the patch for consideration.
--
Joseph S. Myers
[EMAIL PROTECTED]
On Mon, 5 Mar 2007, Ian Lance Taylor wrote:
> I only made one of the changes Jakub suggested: I fixed a typo in the
> documentation. So I didn't bother sending another copy.
>
> For clarity, here is the current version of the patch.
This patch is OK.
--
Joseph S. Myers
[EMAIL PROTECTED]
macros to match
existing and files. Defaults for those macros
will be tricky (for example, glibc and newlib use different choices of
types) and testcases will be needed to make sure the types are the same as
in existing system headers.
--
Joseph S. Myers
[EMAIL PROTECTED]
conf support is used in
any host directory where you wish to use intptr_t / uintptr_t.
Testcases in the GCC testsuite should generally use __SIZE_TYPE__ and
__PTRDIFF_TYPE__ for such cases.
Target libraries such as libstdc++ can probably use size_t and ptrdiff_t
reasonably safely for this.
--
Joseph S. Myers
[EMAIL PROTECTED]
; but if these types ever appear in C++ ABIs,
then you need to care. For purely internal use not appearing in ABIs,
size_t and ptrdiff_t should work.
--
Joseph S. Myers
[EMAIL PROTECTED]
ngconventions.html), to use 1.5 while
avoiding regressions would require auditing every local change since
libtool 1.5 branched to see if it's in 1.5.22.
--
Joseph S. Myers
[EMAIL PROTECTED]
incremental stage should
reduce memory usage and be useful on its own.
--
Joseph S. Myers
[EMAIL PROTECTED]
nt we may not want to XFAIL them - but it's
also true that this FAIL is on 4.2 branch and 4.2.0 is likely to be
released with it. And users installing GCC on common platforms should be
able to use the testsuite to tell whether their build is working OK, which
means that releases should ha
xpected unexpected
FAILs" is helpful neither to users wishing to know if their compiler built
as expected nor to developers glancing over test results to see if they
seem OK.
--
Joseph S. Myers
[EMAIL PROTECTED]
now looking at this (or at a possibly related bootstrap comparison
failure seen on i686-pc-linux-gnu).
--
Joseph S. Myers
[EMAIL PROTECTED]
ve found another issue that can cause comparison failures in some
circumstances; I'll file a PR for that.
--
Joseph S. Myers
[EMAIL PROTECTED]
nge to require that option, which might be simpler.
--
Joseph S. Myers
[EMAIL PROTECTED]
(TYPE_UNSIGNED (node)
+ ? "");
+ }
else
pp_string (buffer, "");
}
--
Joseph S. Myers
[EMAIL PROTECTED]
rees and I expected them all to wish to
use a bug database or instructions URL not containing '@' as the default.
--
Joseph S. Myers
[EMAIL PROTECTED]
he
use of -Wformat-contains-nul being diagnosed just as other such "ignored
without" diagnostics are tested in gcc.dg/format/opt-*.c).
--
Joseph S. Myers
[EMAIL PROTECTED]
On Sun, 25 Mar 2007, Mark Mitchell wrote:
> Joseph S. Myers wrote:
> > On Sat, 24 Mar 2007, Bruce Korb wrote:
> >
> >> This bootstraps in Linux i686 & I can use -Wno-format-contains-nul to
> >> suppress that warning. OK?
> >
> > This is
ause of the rule that a preprocessing token is always the
maximum sequence of characters forming a valid preprocessing token.
--
Joseph S. Myers
[EMAIL PROTECTED]
rror with -pedantic-errors.
In C++, a pedwarn is an error by default, a warning with -fpermissive.
This is completely independent of whether a particular pedwarn call in the
source is executed at runtime. Some are unconditional, and some only
occur if -pedantic.
--
Joseph S. Myers
[EMAIL PROTECTED]
this? Why not use structure
> of bitfields instead of int for target_flags?
You'd need to change everything that uses MASK_FOO|MASK_BAR.
You can specify Var together with Mask in .opt files; that allows you to
create a second variable for flag bits as a smaller patch for now.
--
Joseph S. Myers
[EMAIL PROTECTED]
ically. If existing code were made to use such macros
incrementally, this would also allow a much smaller patch to be used for
any subsequent change to use bit-fields.
--
Joseph S. Myers
[EMAIL PROTECTED]
On Thu, 12 Apr 2007, H. J. Lu wrote:
> Are there any documents/examples for a second variable for flag bits?
config/linux.opt uses Mask and InverseMask with Var.
--
Joseph S. Myers
[EMAIL PROTECTED]
On Thu, 12 Apr 2007, H. J. Lu wrote:
> On Thu, Apr 12, 2007 at 07:03:34PM +0000, Joseph S. Myers wrote:
> > On Thu, 12 Apr 2007, H. J. Lu wrote:
> >
> > > Are there any documents/examples for a second variable for flag bits?
> >
> > config/linux.opt uses Mask
wording to put in the license
text on the source files to allow that.
--
Joseph S. Myers
[EMAIL PROTECTED]
ncy with -static-libgcc.
--
Joseph S. Myers
[EMAIL PROTECTED]
#x27;d need either to use the common diagnostics
infrastructure or add the feature to the Fortran-specific diagnostics
code.
--
Joseph S. Myers
[EMAIL PROTECTED]
e
behavior is undefined.
This is intended to be consistent with all the analysis in N1226.
--
Joseph S. Myers
[EMAIL PROTECTED]
d be necessary to unify two incompatible types from
the same translation unit (even if they'd be compatible in different
translation units), nor to split the uses of a single type within a
translation unit into two or more distinct and incompatible types.
--
Joseph S. Myers
[EMAIL PROTECTED]
lved by taking the trunk version).
--
Joseph S. Myers
[EMAIL PROTECTED]
e. under
GCC_4.3.0. (If it's needed for 4.2 then under 4.2.0 or 4.2.1 depending on
when it goes in 4.2 branch.)
--
Joseph S. Myers
[EMAIL PROTECTED]
t be
any problems like those mentioned in comments in bugzilla-checkout and
htdocs-checkout and cgibin-checkout with the new scripts.
Please ensure that the new script ends up version controlled, whether in
maintainer-scripts in GCC SVN (maybe preferable) or in sourceware
infra/bin CVS. I don&#x
ks", is
live to need conversion to SVN rather than just staying available
read-only from the old repository.)
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
r languages where there is a
common pattern used).
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
hout string pasting capabilities
> if it helps keeping the interface between cpp and the C frontend
> cleaner.
How about calling decl_attributes from fname_decl so a target
insert_attributes hook can add attributes to __func__? Would that suffice
to solve your problem?
--
Joseph S. Mye
uct {int a; int b;} z = PMEM {5312, 3421};
This syntax makes even less sense. A brace-enclosed initializer is not an
object! If z has static storage duration, put the attribute on z. If it
doesn't, how it is initialized is a matter of compiler optimization and
specifying attributes o
iling lists could continue to operate on gcc.gnu.org
(but with posting by nonsubscribers permitted).
Patches need documentation to be updated if applicable - but GPC already
has documentation which I presume is being kept up to date.
--
Joseph S. Myers http://www.srcf.ucam.o
itialization strategy for that variable (NB an
attribute on the variable, using existing syntax, not one on the
initializer itself) - I just don't think it would be useful to do so.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
gi), then just
maintain multiple versions of that small piece of code. This seems to
work well for Ada.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
.def) in gimplification. People are working on more precise
documentation of exactly what the semantics of GENERIC and GIMPLE tree
codes are.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
standing all the standard requirements and the
interaction with GNU extensions and auditing what is currently implemented
and preparing thorough testcases.
void f(int a, int (*b)[*], int c[static sizeof(*b)]);
is valid C99 code. What exactly it means is more obscure.
--
Joseph S. Myers
e with
earlier versions would be substantial. In general, compatibility with the
most recent release series should suffice; if the 4.0 series is
insufficiently stable, effort would better be devoted to improving it than
to keeping compatibility with older and less-maintained series.
--
J
libjava/ChangeLog-2001 has entries from 2000, libstdc++-v3/ChangeLog-2000
has entries from 1998 and 1999. Perhaps a followup patch could also fix
those?
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSour
eginning be using the mirror rather than
having a lot on gcc.gnu.org as you would if mirroring was only set up some
time after initial announcement).
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
erhaps a note to gcc-announce that a move is proposed and
people who use anoncvs or mirror the CVS tree might like to make sure they
are ready (in particular, check whether subversion works on their
platforms) would be a good idea well in advance.
--
Joseph S. Myers http://www.sr
for generating diffs it won't include the content of a file being copied
with no changes, so it should be safe to configure it to send diffs for
all of add, copy, modify and delete without generating excessively large
diffs for routine tag/branch operations.)
--
Joseph S. Mye
erface with tree-ssa, and as soon as it goes in GCC CVS it's
inevitable changes tested only with GCC CVS will happen to break support
for some previous back end.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PR
eph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
ffs and attaching diffs for non-UTF-8
files (some message catalogs and testcases) to avoid problems with
messages claiming to be in UTF-8 which aren't.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSou
precision,
precision, length modifier, scanf 'a' kludge, conversion type specifier)
out of the code and into the datastructures.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
re carried out before folding, fold doesn't
need to concern itself with preserving language-specific properties of the
trees.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla assignments and CCs)
end maintainers. As extensions go, it's actually not
> that bad; the semantics are relatively well defined.
The min/max expression extension is C++ only, the C front end doesn't have
it.
(Removing the extension is separate from the question of whether MIN_EXPR
and MAX_EXPR are useful
ontents of our website, or get more reviewers. For instance, we could think
> of a policy where a www patch can be applied after 48hrs if nobody says
> otherwise).
You may not have noticed that Gerald is away until 13 March. Otherwise
website patches do get reviewed quickly.
--
Joseph
-number }
or
{ dg-error "expected regexp" "test name" { xfail *-*-* } line-number }
> The general frustration is: where is dg-error documented?
It ought to be in the dejagnu manual (i.e., that's where documentation
should best be contributed) since dg-error is part of ba
The pre-Lillehammer WG14 mailing includes N
<http://www.open-std.org/jtc1/sc22/wg14/www/docs/n.htm> with an
updated analysis of the DR#236 aliasing issues taking account of comments
previously made on this list.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm
used in the glibc dynamic linker. I'm not sure why,
and they may be inline nested functions whose addresses are never taken.
The extension is not present in GNU C++, only in GNU C.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
inter-to-float should then result in a
plain pointer-to-float.
We don't yet implement either attributes on array parameters applying to
the array not the pointer, or attributes inside the [] of the array
parameter applying to the pointer. (This is documented in "Attri
any qualifiers, to apply attributes to the pointer rather than to the
array, (b) attributes on the parameter declarations that should apply to
the array instead of to the pointer (instead they'll just apply to the
pointer to which the array decays).
--
Joseph S. Myers http:
to int in arithmetic. Bit-field types don't have their own
modes, but in principle you should be able to have a special type with its
own mode narrower than char: however, you may need to implement
optimizations which convert operations on promoted types to operations on
narrow types for
g mistaken for dead because of insufficient development
activity on the GCC lists, keeping testresults going to gcc-testresults
provides the evidence that the target is not dead after all.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mai
s ice-on-valid; the construct in
> question has undefined behaviour.
Undefined behavior on execution, not on translation. Undefined behavior
only if bar is called and the call to f from bar returns.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (
*doesn't* warn for
those constructs. And, indeed, the test passes: GCC does not warn for
these constructs.
If you have a genuine bug in -Wsequence-point, report it to GCC Bugzilla
with a complete testcase to reproduce the bug.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
hey exceed the message size limit, and
this remains the case for a prolonged period such as ever since 4.0
branched, that also indicates lack of active maintenance.)
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PRO
her testing the old compiler with the new
testsuite (e.g. using contrib/test_installed, but that doesn't do
libstdc++ testing), or running the tests manually with the old compiler.
(Whether or not it is a regression, it is still worth a bug report in
Bugzilla, as is any FAIL or XFAIL.)
--
tch for which the branch is still open).
In checking for any similar references which should be fixed I noticed
that config/arm/libgcc-bpabi.ver defines a GCC_3.5 symbol version, but
it's probably rather too late to change that and may not be desirable to
change it anyway.
ing CSE and constant folding
which might be affected by changes in rounding mode. Certainly some of
Stephen Moshier's testcases (attached to bug 20785) fail.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTE
ribute. (See how this is done
for vector_size attributes.)
The __section__ attribute is purely a storage-class-like attribute. If
you want DTR18037 address space specifiers, feel free to implement them;
they are probably the least problematic part of that TR.
--
Joseph S. Myers
t;
> The flag that controls this is -ftrapping-math, and it defaults to "on".
I was thinking of -frounding-math, which defaults to "off" and comes with
a warning in the documentation that it may not yet do everything required.
--
Joseph S. Myers http://www.src
d take
care to avoid getting GNU iconv.h but libc iconv or GNU libiconv but libc
iconv.h (the include and linker search paths used in configuration may not
be consistent if --with-libiconv-prefix hasn't been used).
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL
Results for hppa2.0w-hp-hpux11.11, no regressions:
http://gcc.gnu.org/ml/gcc-testresults/2005-04/msg01379.html
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PROTECTED] (Bugzilla
e a configuration which works for 4.0.0 and mainline, as
well as for csl-sol210-3_4-branch from which /usr/sfw/bin/gcc comes.
--
Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/
[EMAIL PROTECTED] (personal mail)
[EMAIL PROTECTED] (CodeSourcery mail)
[EMAIL PRO
1 - 100 of 1215 matches
Mail list logo
