New gfortran bug
in [Fortran]
|
Prev: Expressive programming for computational physics in Fortran 95+ (paper)
Next: Allocatable versus automatic arrays
From: Al Greynolds on 1 Jun 2010 17:22 Can somebody submit the following (minor) bug in gfortran 4.6 for me? I haven't had much luck with Bugzilla. character(1) c read(*,*) c ! latest 4.6 doesn't like the following unless broken into separate lines if (c=='y') stop; if (c=='Y') stop end Al Greynolds www.ruda-cardinal.com
From: steve on 1 Jun 2010 17:29 On Jun 1, 2:22 pm, Al Greynolds <awgreyno...(a)earthlink.net> wrote: > Can somebody submit the following (minor) bug in gfortran 4.6 for me? > I haven't had much luck with Bugzilla. > > character(1) c > read(*,*) c > ! latest 4.6 doesn't like the following unless broken into separate > lines > if (c=='y') stop; if (c=='Y') stop > end > Can you define latest? It works for me with gfortran 4.6.0 20100510. -- steve
From: Al Greynolds on 1 Jun 2010 17:44 On Jun 1, 2:29 pm, steve <kar...(a)comcast.net> wrote: > On Jun 1, 2:22 pm, Al Greynolds <awgreyno...(a)earthlink.net> wrote: > > > Can somebody submit the following (minor) bug in gfortran 4.6 for me? > > I haven't had much luck with Bugzilla. > > > character(1) c > > read(*,*) c > > ! latest 4.6 doesn't like the following unless broken into separate > > lines > > if (c=='y') stop; if (c=='Y') stop > > end > > Can you define latest? > > It works for me with gfortran 4.6.0 20100510. > > -- > steve I know an equation.com Window's build from 20100515 showed the problem. Al
From: steve on 1 Jun 2010 21:59 On Jun 1, 2:44 pm, Al Greynolds <awgreyno...(a)earthlink.net> wrote: > On Jun 1, 2:29 pm, steve <kar...(a)comcast.net> wrote: > > > > > On Jun 1, 2:22 pm, Al Greynolds <awgreyno...(a)earthlink.net> wrote: > > > > Can somebody submit the following (minor) bug in gfortran 4.6 for me? > > > I haven't had much luck with Bugzilla. > > > > character(1) c > > > read(*,*) c > > > ! latest 4.6 doesn't like the following unless broken into separate > > > lines > > > if (c=='y') stop; if (c=='Y') stop > > > end > > > Can you define latest? > > > It works for me with gfortran 4.6.0 20100510. > > > -- > > steve > > I know an equation.com Window's build from 20100515 showed the > problem. > > Al http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44371 This appears to be regression introduced when STOP and ALL_STOP were updated to F2008 features. I've posted a suggested fix in the PR. -- steve
From: Jerry DeLisle on 2 Jun 2010 00:56
On 06/01/2010 06:59 PM, steve wrote: > On Jun 1, 2:44 pm, Al Greynolds<awgreyno...(a)earthlink.net> wrote: >> On Jun 1, 2:29 pm, steve<kar...(a)comcast.net> wrote: >> >> >> >>> On Jun 1, 2:22 pm, Al Greynolds<awgreyno...(a)earthlink.net> wrote: >> >>>> Can somebody submit the following (minor) bug in gfortran 4.6 for me? >>>> I haven't had much luck with Bugzilla. >> >>>> character(1) c >>>> read(*,*) c >>>> ! latest 4.6 doesn't like the following unless broken into separate >>>> lines >>>> if (c=='y') stop; if (c=='Y') stop >>>> end >> >>> Can you define latest? >> >>> It works for me with gfortran 4.6.0 20100510. >> >>> -- >>> steve >> >> I know an equation.com Window's build from 20100515 showed the >> problem. >> >> Al > > http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44371 > > This appears to be regression introduced when STOP > and ALL_STOP were updated to F2008 features. I've > posted a suggested fix in the PR. > I just committed the fix to 4.6 trunk. Thanks Al for the report. If Bugzilla bugs you, one can post to fortran(a)gcc.gnu.org. Thanks Steve for help with the test cases. Thanks Tobias for submitting the bugzilla so I would see this. My apologies for the breakage. Regards, Jerry |