On Thu, Oct 24, 2013 at 3:04 PM, Glenn Fowler <glenn.s.fow...@gmail.com> wrote:
> On Thu, Oct 24, 2013 at 8:40 AM, Roland Mainz <roland.ma...@nrubsig.org>
> wrote:
>> While trying to build a working ksh93 from ast-ksh.2013-10-10 on SuSE
>> 12.3/AMD/64bit using the PathScale pathcc compiler I hit the following
>> warnings:
>> -- snip --
>> + cc -D_BLD_DLL -fPIC -D_BLD_ast '-DSHOPT_CMDLIB_BLTIN=0'
>> '-DSH_CMDLIB_DIR="/usr/ast/bin"'
>>
>> '-DSHOPT_CMDLIB_HDR="/home/test001/work/ast_ksh_20131010/build_pathcc/tmp_gnulinux_built
>> in_header.h"' -DSHOPT_SYSRC '-D_map_libc=1' -O2 -fstrict-aliasing
>> -Wstrict-aliasing -Wsequence-point -Wno-parentheses -Wno-unused
>> -Wno-trigraphs -Wuninitialized -Waddress -I. -I/h
>> ome/test001/work/ast_ksh_20131010/build_pathcc/src/lib/libast -Icomp
>> -I/home/test001/work/ast_ksh_20131010/build_pathcc/src/lib/libast/comp
>> -Iport -I/home/test001/work/ast_ksh_201
>> 31010/build_pathcc/src/lib/libast/port -Iinclude
>> -I/home/test001/work/ast_ksh_20131010/build_pathcc/src/lib/libast/include
>> -Istd -I/home/test001/work/ast_ksh_20131010/build_pathcc
>> /src/lib/libast/std -D_PACKAGE_ast -c
>>
>> /home/test001/work/ast_ksh_20131010/build_pathcc/src/lib/libast/string/wcstoutf32s.c
>>
>> /home/test001/work/ast_ksh_20131010/build_pathcc/src/lib/libast/string/wcstoutf32s.c:48:34:
>> warning: comparison of unsigned expression < 0 is always false
>> [-Wtautological-compare]
>> if (mbconv(tmp, wchar[i], &q) < 0)
>> ~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~
>>
>> /home/test001/work/ast_ksh_20131010/build_pathcc/src/lib/libast/string/wcstoutf32s.c:64:106:
>> warning: comparison between pointer and integer ('void *' and 'int')
>> if (ast.mb_wc2uc == (void*)(-1) && (ast.mb_wc2uc =
>> (void*)iconv_open("UTF-8", codeset(CODESET_ctype))) == -1)
> thx roland
> this was coded during the rush to fix mb* wc* before losing access to the
> sw/download/
> a last minute change was required to get the ast iconv intercept working
> and the effects of that change were not completely propagated
>
> does this fix the 8Gib tr allocation bug?
AFAIK no... this is only for the \w[]/\u[] codepaths which fail in
various ways if the pathcc compiler optimiser is turned on... I have
to rework the patch a bit again to weed-out another bug in these two
source files... updated patch is currently being tested...
> (again dgk and i are temporarily hobbled by incomplete build envs)
I know... ;-(
----
Bye,
Roland
--
__ . . __
(o.\ \/ /.o) roland.ma...@nrubsig.org
\__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer
/O /==\ O\ TEL +49 641 3992797
(;O/ \/ \O;)
Email secured by Check Point
_______________________________________________
ast-developers mailing list
ast-developers@lists.research.att.com
http://lists.research.att.com/mailman/listinfo/ast-developers
