[ https://issues.apache.org/jira/browse/STDCXX-237?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Martin Sebor updated STDCXX-237: -------------------------------- Severity: Runtime Error Fix Version/s: (was: 4.2) 5.0 Scheduled for 5.0 since, like STDCXX-448, the fix is binary incompatible. > SIGFPE in std::operator%(slice, slice) > -------------------------------------- > > Key: STDCXX-237 > URL: https://issues.apache.org/jira/browse/STDCXX-237 > Project: C++ Standard Library > Issue Type: Bug > Components: 26. Numerics > Affects Versions: 4.1.2, 4.1.3 > Reporter: Martin Sebor > Assignee: Martin Sebor > Fix For: 5.0 > > > Moved from the Rogue Wave bug tracking database: > ****Created By: sebor @ Apr 27, 2000 10:43:55 PM**** > Subject: valarray/gslice problem > Date: Fri, 21 Apr 2000 04:18:09 -0700 > From: Dennis Handly > To: sebor@roguewave.com > CC: dhandly@hpcllmx.cup.hp.com, wanc@cup.hp.com > The following test fails with a SIGFPE (% by 0): > #include > int main() { > std::valarray ival; > std::gslice test_gslice; > std::operator%(ival[test_gslice], ival[test_gslice]); > } > test_gslice should select no elements. > I think the problem is in : > inline size_t gslice::ind_numb() const > { > size_t value = 1; > for(size_t ind=0; ind< length_.size(); ind++ ) > value*= length_[ind]; > return value; > } > This doesn't return 0 if length_.size() is 0. > ****Modified By: sebor @ May 18, 2000 12:01:47 PM**** > Subject: Re: valarray/gslice problem > Date: Thu, 11 May 2000 22:14:58 -0700 > From: Dennis Handly > To: sebor@roguewave.com > CC: dhandly@hpcllmx.cup.hp.com, wanc@cup.hp.com > >The incident number is 22770. > I changed valarray to check for 0 and return 0: > inline size_t gslice::ind_numb() const > { > size_t value = 1; > register size_t lsize = _C_length.size(); > if (lsize == 0) > return 0; > for(size_t __i=0; __i < lsize; __i++ ) -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.