[prev in list] [next in list] [prev in thread] [next in thread]
List: pcc-list
Subject: Re: pcc-2010-09-14: internal compiler error on GNU/Linux AMD64
From: Anders Magnusson <ragge () ludd ! ltu ! se>
Date: 2010-09-20 17:03:54
Message-ID: 4C9793FA.5080903 () ludd ! ltu ! se
[Download RAW message or body]
Now long double should work.
-- Ragge
On 09/15/2010 03:11 AM, Nelson H. F. Beebe wrote:
> The following test program (extracted from a much larger real program)
> compiles on x86 (IA-32) with the pcc-2010-09-14 nightly snapshot, but
> fails on x86_64 (AMD64, EM64T):
>
> % cat bug005.c
> extern void __assert_fail (__const char *__assertion,
> __const char *__file,
> unsigned int __line,
> __const char *__function)
> __attribute__ ((__nothrow__))
> __attribute__ ((__noreturn__));
>
> void
> bug003(double zz)
> {
> ((((-2.1895e-2L)< zz)&& (zz< (2.1895e-2L)))
> ? (void) (0)
> : (__assert_fail ("((-2.1895e-2L)< zz)&& (zz< (2.1895e-2L))",
> "foo.h", 222, __PRETTY_FUNCTION__),
> (void) (0)));
> }
>
> % pcc -v
> pcc 0.9.9 for x86_64-unknown-linux-gnu, beebe@XXXXXXXX.math.utah.edu Tue Sep 14
> 12:48:49 MDT 2010
>
> % pcc -c bug005.c
> bug005.c, line 18: compiler error: Cannot generate code, node 0x59a70a0 op>=
>
> If the L suffixes are removed from the constants, the code compiles
> successfully. However, the L suffixes do not precipitate warnings
> about "long double not yet implemented."
>
> Notice that no header files are needed. The error persists if the
> function prototype for __assert_fail() is removed, so it should be
> relatively easy to track down the source of the error inside the
> compiler.
>
> -------------------------------------------------------------------------------
> - Nelson H. F. Beebe Tel: +1 801 581 5254 -
> - University of Utah FAX: +1 801 581 4148 -
> - Department of Mathematics, 110 LCB Internet e-mail: beebe@math.utah.edu -
> - 155 S 1400 E RM 233 beebe@acm.org beebe@computer.org -
> - Salt Lake City, UT 84112-0090, USA URL: http://www.math.utah.edu/~beebe/ -
> -------------------------------------------------------------------------------
>
>
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic