This gcc version always requires that system have at least one
floating value that is usable as `NAN' - in its own headers or
specified in build configuration. In this system with this compiler
it is not the case. Can not obtain `NAN' value at all. When trying
to compute it dynamically, the process just gets <Floating point
exception> signal, so can not even obtain the value in it and copy
into `NAN' definition.
Environment:
System: OSF1 . V5.1 2650 alpha
Machine:
Using native `CC=cc', `CFLAGS='-pthread -g''.
host: alpha-dec-osf5.1b
build: alpha-dec-osf5.1b
target: alpha-dec-osf5.1b
configured with: ../gcc-4.0.2/configure --disable-nls --disable-libgcj
--enable-languages=c,c++,objc
How-To-Repeat:
gmake[1]: Entering directory `gcc-4.0.2-e/build-alpha-dec-osf5.1b/libiberty'
make bootstrap
cc -c -DHAVE_CONFIG_H -pthread -g -I.
-I../../../gcc-4.0.2/libiberty/../include
../../../gcc-4.0.2/libiberty/floatformat.c -o floatformat.o
cc: Error: ../../../gcc-4.0.2/libiberty/floatformat.c, line 319: In this
statement, the libraries on this platform do not yet support compile-time
evaluation of the constant expression "0.0/0.0". (constfoldns)
dto = NAN;
--------------^
gmake[1]: *** [floatformat.o] Error 1
The preprocessed line 319 is as follows.
dto = ( 0.0 / 0.0 ) ;
------- Comment #1 from gin at mo dot msk dot ru 2005-11-23 15:02 -------
Fix:
Do not have patch currently. Most likely gcc has to check whether
`NAN' value exists, and compile some code in `floatformat.c' only if
it is the case. Or at least allow to disable compiling that code in
build configuration.
--
Summary: build breaks if no `NAN'
Product: gcc
Version: 4.0.2
Status: UNCONFIRMED
Severity: critical
Priority: P2
Component: bootstrap
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: gin at mo dot msk dot ru
GCC build triplet: alpha-dec-osf5.1b
GCC host triplet: alpha-dec-osf5.1b
GCC target triplet: alpha-dec-osf5.1b
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=25002
