Re: CVS commit: src/crypto/dist/openssl/apps

2009-04-14 Thread Robert Elz
Date:Wed, 15 Apr 2009 00:08:30 +0100 From:David Laight Message-ID: <20090414230830.gc4...@snowdrop.l8s.co.uk> | The ctype functions are often implemented as #defines that index arrays, | so if the argument is 'char' it is very likely to index off the front | whi

Re: CVS commit: src/crypto/dist/openssl/apps

2009-04-14 Thread David Laight
On Wed, Apr 15, 2009 at 01:38:02AM +0700, Robert Elz wrote: > > But since both you, and David, have commented on that issue, I took > a look, and I certainly would not agree with "simply wrong". I might > agree with "simply useless", after all, ctyoe.h has > > int toupper(int); > > so

Re: CVS commit: src/usr.bin/w

2009-04-14 Thread Luke Mewburn
On Tue, Apr 14, 2009 at 07:37:32AM +, Luke Mewburn wrote: | Module Name:src | Committed By: lukem | Date: Tue Apr 14 07:37:32 UTC 2009 | | Modified Files: | src/usr.bin/w: w.c Previously empty log message updated to Fix WARNS=4 issues |

Re: CVS commit: src/crypto/dist/openssl/apps

2009-04-14 Thread Robert Elz
Date:Tue, 14 Apr 2009 19:59:46 +0200 From:Joerg Sonnenberger Message-ID: <20090414175946.ga12...@britannica.bec.de> | toupper like the rest of ctype.h is defined for EOF and for all values | of a unsigned char. Yes. Which is why its arg cannot be u_char as you

Re: CVS commit: src/crypto/dist/openssl/apps

2009-04-14 Thread Joerg Sonnenberger
On Wed, Apr 15, 2009 at 12:40:26AM +0700, Robert Elz wrote: > If EOF is to be a valid parameter to toupper(), then the arg type > cannot be u_char - however much you'd like it to be. toupper like the rest of ctype.h is defined for EOF and for all values of a unsigned char. unsigned char is by defi

Re: CVS commit: src/crypto/dist/openssl/apps

2009-04-14 Thread David Laight
On Wed, Apr 15, 2009 at 12:40:26AM +0700, Robert Elz wrote: > Date:Tue, 14 Apr 2009 16:11:13 +0200 > From:Joerg Sonnenberger > Message-ID: <20090414141113.gd...@britannica.bec.de> > > | toupper() gets u_char arguments, not int. This cast is simply wrong. > > Accord

Re: CVS commit: src/crypto/dist/openssl/apps

2009-04-14 Thread Robert Elz
Date:Tue, 14 Apr 2009 16:11:13 +0200 From:Joerg Sonnenberger Message-ID: <20090414141113.gd...@britannica.bec.de> | toupper() gets u_char arguments, not int. This cast is simply wrong. According to TFM ... SYNOPSIS #include int toupper(int c); an

Re: CVS commit: src/crypto/dist/openssl/apps

2009-04-14 Thread Joerg Sonnenberger
On Tue, Apr 14, 2009 at 10:03:55AM +, Luke Mewburn wrote: > Module Name: src > Committed By: lukem > Date: Tue Apr 14 10:03:55 UTC 2009 > > Modified Files: > src/crypto/dist/openssl/apps: ca.c > > Log Message: > Call toupper() with an int argument. toupper() gets u_char argume

Re: CVS commit: src/common/lib/libprop

2009-04-14 Thread Joerg Sonnenberger
On Tue, Apr 14, 2009 at 02:53:41AM +, Adam Hamsik wrote: > Log Message: > Check if pd is not NULL before we try to lock rw lock associated with it. > This fixes proplib crash when NULL is passed to prop_dictionary_get as a > dictionary. Why is that valid operation? Joerg

Re: CVS commit: src/compat/external/bsd/bind/lib

2009-04-14 Thread Martin Husemann
On Tue, Apr 14, 2009 at 10:39:38AM +, Martin Husemann wrote: > Module Name: src > Committed By: martin > Date: Tue Apr 14 10:39:38 UTC 2009 > > Modified Files: > src/compat/external/bsd/bind/lib/libbind9: Makefile > src/compat/external/bsd/bind/lib/libdns: Makefile >

Re: CVS commit: src/usr.bin/w

2009-04-14 Thread Thomas Klausner
On Tue, Apr 14, 2009 at 07:37:32AM +, Luke Mewburn wrote: > Module Name: src > Committed By: lukem > Date: Tue Apr 14 07:37:32 UTC 2009 > > Modified Files: > src/usr.bin/w: w.c > > > To generate a diff of this commit: > cvs rdiff -u -r1.73 -r1.74 src/usr.bin/w/w.c > > Please