From: UnixOS2 Archive To: "UnixOS2 Archive" Date: Sat, 22 Feb 2003 04:53:55 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600 Subject: [UnixOS2_Archive] No. 52 ************************************************** Friday 21 February 2003 Number 52 ************************************************** Subjects for today 1 Re: Building NCURSES using Posix/2 : Stefan Neis 2 Re: Building NCURSES using Posix/2 : Thomas Dickey 3 Re: Building NCURSES using Posix/2 : Thomas Dickey 4 Porting Unix USB apps to OS/2 : John Poltorak 5 Re: Building NCURSES using Posix/2 : John Poltorak 6 Re: Porting Unix USB apps to OS/2 : Bart van Leeuwen" 7 Re: Building NCURSES using Posix/2 : John Poltorak 8 Re: Building NCURSES using Posix/2 : John Poltorak 9 Re: Porting Unix USB apps to OS/2 : Sebastian Wittmeier (ShadoW)" 10 Re: Building NCURSES using Posix/2 : Stefan Neis **= Email 1 ==========================** Date: Sat, 22 Feb 2003 00:15:35 +0100 (CET) From: Stefan Neis Subject: Re: Building NCURSES using Posix/2 On Fri, 21 Feb 2003, John Poltorak wrote: > Would your patch have made any difference in this case? Probably. At least all these functions that it claims to be implicitly declared are declared in netdb.h, which it probably didn't find without the patch. Anyway, I ended up with a more or less complete build, but it's having issues with malloc, which I don't quite understand ... While we are at patching: os2\os2.c also needs some fixes as well, as it unconditionally defines a couple of functions that are already provided ny Posix/2, so if perl adds its own (dummy) implementations for those, the linker complains about multiply defined symbols... Wasn't there a OS/2 specific fix with respect to malloc/Perl_malloc? Anyone knows details about the changed file(s)? I assume it fails to handle Posix/2 correctly. Regards, Stefan -- Micro$oft is not an answer. It is a question. The answer is 'no'. **= Email 2 ==========================** Date: Sat, 22 Feb 2003 10:48:07 -0500 From: Thomas Dickey Subject: Re: Building NCURSES using Posix/2 On Sat, Feb 22, 2003 at 03:33:19PM +0000, John Poltorak wrote: > I get Hunk #1 FAILED at 5032. > > How do I tell why? there should be a ".rej" file which consists of the rejected chunks. -- Thomas E. Dickey http://invisible-island.net ftp://invisible-island.net **= Email 3 ==========================** Date: Sat, 22 Feb 2003 11:30:15 -0500 From: Thomas Dickey Subject: Re: Building NCURSES using Posix/2 On Sat, Feb 22, 2003 at 03:59:50PM +0000, John Poltorak wrote: > On Sat, Feb 22, 2003 at 10:48:07AM -0500, Thomas Dickey wrote: > > On Sat, Feb 22, 2003 at 03:33:19PM +0000, John Poltorak wrote: > > > > > I get Hunk #1 FAILED at 5032. > > > > > > How do I tell why? > > > > there should be a ".rej" file which consists of the rejected chunks. > > Yes, there is but I can't identify what is wrong. but you're missing a piece - the .rej file only copies part of the patch. It would only be applied if the context (lines before/after the change) matched. The mismatch may be as simple as a trailing blank on one of the lines. Here are the actual lines from perl 5.8.0: echo "#include <\$wanted>" > foo\$\$.c $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ $cppfilter $grep "^[ ]*#.*\$wanted" | \ while read cline; do name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` case "\$name" in Note that the last two have tabs rather than spaces. You can make patch accept that by using the "-l" option. > Here is the patch:- > > --- perl-5.8.0-org/Configure Fri Jan 31 22:02:22 2003 > +++ perl-5.8.0/Configure Sun Feb 2 18:48:52 2003 > at at -5032,6 +5032,7 at at > echo "#include <\$wanted>" > foo\$\$.c > $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ > $cppfilter $grep "^[ ]*#.*\$wanted" | \ > +sed "s/\\\\\\\\\\\\\\\\/\//g"| \ > while read cline; do > name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` > case "\$name" in > > > > Here is Configure.rej:- > > *************** > *** 5032,5037 **** > echo "#include <\$wanted>" > foo\$\$.c > $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ > $cppfilter $grep "^[ ]*#.*\$wanted" | \ > while read cline; do > name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` > case "\$name" in > --- 5032,5038 ---- > echo "#include <\$wanted>" > foo\$\$.c > $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ > $cppfilter $grep "^[ ]*#.*\$wanted" | \ > + sed "s/\\\\\\\\\\\\\\\\/\//g"| \ > while read cline; do > name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` > case "\$name" in > > > I don't see anything wrong. Could it be a bug in patch? > > > > > -- > > Thomas E. Dickey > > http://invisible-island.net > > ftp://invisible-island.net > > > -- > John > -- Thomas E. Dickey http://invisible-island.net ftp://invisible-island.net **= Email 4 ==========================** Date: Sat, 22 Feb 2003 11:42:23 +0000 From: John Poltorak Subject: Porting Unix USB apps to OS/2 Does anyone know what is involved in porting Unix apps which handle a USB interface to OS/2? The current OS/2 port of PilotLink is missing USB functionality, but it is available in the original source. -- John **= Email 5 ==========================** Date: Sat, 22 Feb 2003 15:33:19 +0000 From: John Poltorak Subject: Re: Building NCURSES using Posix/2 On Fri, Feb 21, 2003 at 10:00:05PM +0100, Stefan Neis wrote: > On Fri, 21 Feb 2003, John Poltorak wrote: > > > Wonder if I should try it out on Perl yet... > > Well, it might even work. Note however that you'll still need the > appended patch to Perl's Configure to get rid of that problem with > netdb.h not being found (In human readable language, that's just > "replace \\ by /", it only looks so complicated because of all the > escaping involved ...). > > Regards, > Stefan > --------------------- Patch --------- > --- Configure~ Fri Jan 31 22:02:22 2003 > +++ Configure Sun Feb 2 18:48:52 2003 > at at -5032,6 +5032,7 at at > echo "#include <\$wanted>" > foo\$\$.c > $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ > $cppfilter $grep "^[ ]*#.*\$wanted" | \ > +sed "s/\\\\\\\\\\\\\\\\/\//g"| \ > while read cline; do > name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` > case "\$name" in > I get Hunk #1 FAILED at 5032. How do I tell why? -- John **= Email 6 ==========================** Date: Sat, 22 Feb 2003 15:49:41 +0100 From: "Bart van Leeuwen" Subject: Re: Porting Unix USB apps to OS/2 On 22-02-2003 12:42:23 owner-os2-unix wrote: >Does anyone know what is involved in porting Unix apps which handle a USB >interface to OS/2? > >The current OS/2 port of PilotLink is missing USB functionality, but it is >available in the original source. The netlabs usb project provides a driver which sort of mimics the libusb interface. but if programs use other means off addressing the USB bus it will be a hard task I think This is the same approach I took with gphoto2 usb support Bart **= Email 7 ==========================** Date: Sat, 22 Feb 2003 15:59:50 +0000 From: John Poltorak Subject: Re: Building NCURSES using Posix/2 On Sat, Feb 22, 2003 at 10:48:07AM -0500, Thomas Dickey wrote: > On Sat, Feb 22, 2003 at 03:33:19PM +0000, John Poltorak wrote: > > > I get Hunk #1 FAILED at 5032. > > > > How do I tell why? > > there should be a ".rej" file which consists of the rejected chunks. Yes, there is but I can't identify what is wrong. Here is the patch:- --- perl-5.8.0-org/Configure Fri Jan 31 22:02:22 2003 +++ perl-5.8.0/Configure Sun Feb 2 18:48:52 2003 at at -5032,6 +5032,7 at at echo "#include <\$wanted>" > foo\$\$.c $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ $cppfilter $grep "^[ ]*#.*\$wanted" | \ +sed "s/\\\\\\\\\\\\\\\\/\//g"| \ while read cline; do name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` case "\$name" in Here is Configure.rej:- *************** *** 5032,5037 **** echo "#include <\$wanted>" > foo\$\$.c $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ $cppfilter $grep "^[ ]*#.*\$wanted" | \ while read cline; do name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` case "\$name" in --- 5032,5038 ---- echo "#include <\$wanted>" > foo\$\$.c $cppstdin $cppminus $cppflags < foo\$\$.c 2>/dev/null | \ $cppfilter $grep "^[ ]*#.*\$wanted" | \ + sed "s/\\\\\\\\\\\\\\\\/\//g"| \ while read cline; do name=\`echo \$cline | $awk "\$awkprg" | $tr -d '"'\` case "\$name" in I don't see anything wrong. Could it be a bug in patch? > -- > Thomas E. Dickey > http://invisible-island.net > ftp://invisible-island.net -- John **= Email 8 ==========================** Date: Sat, 22 Feb 2003 16:27:58 +0000 From: John Poltorak Subject: Re: Building NCURSES using Posix/2 On Sat, Feb 22, 2003 at 03:59:50PM +0000, John Poltorak wrote: > On Sat, Feb 22, 2003 at 10:48:07AM -0500, Thomas Dickey wrote: > > On Sat, Feb 22, 2003 at 03:33:19PM +0000, John Poltorak wrote: > > > > > I get Hunk #1 FAILED at 5032. > > > > > > How do I tell why? > > > > there should be a ".rej" file which consists of the rejected chunks. > > Yes, there is but I can't identify what is wrong. > I don't see anything wrong. After recreating the patch, it looks to be related to tabs being translated to spaces - always a potential problem when it comes to cutting and pasting... > > > -- > > Thomas E. Dickey > > http://invisible-island.net > > ftp://invisible-island.net -- John **= Email 9 ==========================** Date: Sat, 22 Feb 2003 17:37:21 +0100 (CET) From: "Sebastian Wittmeier (ShadoW)" Subject: Re: Porting Unix USB apps to OS/2 On Sat, 22 Feb 2003 15:49:41 +0100, Bart van Leeuwen wrote: >The netlabs usb project provides a driver which sort of mimics the libusb >interface. PilotLink uses the "visor.o" kernel module. >but if programs use other means off addressing the USB bus it will be a >hard task I think In earlier versions the task of that module was handled by the standard usb serial module. So it shouldn't be too difficult in that case. >This is the same approach I took with gphoto2 usb support Would be nice, if you give us advice based on your experience. Sebastian **= Email 10 ==========================** Date: Sat, 22 Feb 2003 18:54:05 +0100 (CET) From: Stefan Neis Subject: Re: Building NCURSES using Posix/2 On Sat, 22 Feb 2003, John Poltorak wrote: > After recreating the patch, it looks to be related to tabs being > translated to spaces - always a potential problem when it comes to > cutting and pasting... Exactly. :-( I actually wasn't aware of that problem as everything for which I submitted patches so far is living on a "spaces only" policy (except, of course, for Makefiles...). Again, I learned something... :-) Regards, Stefan -- Micro$oft is not an answer. It is a question. The answer is 'no'.