I've got a partial patch which works with older (4.3) gccs, but fails
gimple's check for trunk (attached). My trivial test case...
char *
foo (char *a, int b)
{
return a-b;
}
...fails thusly:
constant 32>
unit size constant 4>
align 8 symtab 0 alias set -1 canonical type 0xb7f52c
We are starting to get some of the basic WHOPR functionality
implemented, so maybe we should think about offering different flags
for it. Currently, we have these flags:
-flto:
Loads all the bodies from all the files given on the command line and
optimizes everything in memory.
-flto-single
Load
Hi,
At first, I'm not sure it is right place to discussing this.
If not, please kindly let me know where the right place is.
On section 10.7 Constant Expression Types,
4th, 5th, 6th and 7th paragraphs of "const_vector" description seem to be
misplaced.
They seem to be applied to "CONST_DOUBLE_M
The perspective commercial LTD seeks for new members
If you possess 3 free hours every week, a small experience in computers and
free phone to which we can call you, you have possibility to start work with
us and have more than 2000 US dollars
If you are interested in our proposition
On Mon, Sep 29, 2008 at 06:23:13PM -0400, Robert Dewar wrote:
> Joe Buck wrote:
> >On Mon, Sep 29, 2008 at 02:54:26PM +0100, Joern Rennecke wrote:
> >>Couldn't the plugin interface require a copyrighted passphrase to be
> >>sent by the plugin when it registeres, and the passphrase can then
> >>be l
I'm defining some peepholes for my machine. But I've
got some troubles :
I want to use peephole to reduce :
ADD R4, 1
CMP R4, a--> CMPD1 R4, a - 1.
The assembly code that cc1.exe generates has two
instruction :
ADD R4, 1
CMP R4, a
But when I define peepholes to reduce it, t
I'm defining some peepholes for my machine. But I've
got some troubles :
I want to use peephole to reduce :
ADD R4, 1
CMP R4, a--> CMPD1 R4, a - 1.
The assembly code that cc1.exe generates has two
instruction :
ADD R4, 1
CMP R4, a
But when I define peepholes to reduce it, t
Dong Phuong wrote on 30 September 2008 17:32:
> But when I define peepholes to reduce it, there's
> nothing change. When I look at the rtl expression, I
> see that there something beetween the instruction ADD
> and CMP :
>
> (insn 18 16 19 0x0 (set (reg/v:HI 22)
> (plus:HI (reg/v:HI 22)
Andreas Tobler wrote:
Hi Peter,
Peter O'Gorman wrote:
Andreas Tobler wrote:
Jack Howarth wrote:
On i686-apple-darwin9, I am seeing massive regressions in the
libjava
testsuite in revision 140713 compared to my previous test on 20080925.
Since the libtool updates went in (which would see t
