headers
Roland Illig | 1 Dec 2005 01:52
Picon
Picon

Re: proposal of how to deal with missing header files supplied by nbcompat

Georg Schwarz wrote:
> In Addition, PLIST would have an extra entry
> 
> ${NEED_REGEX_H}include/regex.h
> 
> libnbcompat's Makefile would finally need
> 
> .if defined(NEED_REGEX_H) && !empty(NEED_REGEX_H:M[Yy][Ee][Ss])
> PLIST_SUBST+=   NEED_REGEX_H=
> .else
> PLIST_SUBST+=   NEED_REGEX_H=" <at> comment "
> .endif

It's not specific to this thread, but I would like the PLIST "variables" 
to have a certain naming convention. NEED_REGEX_H is fine for Makefiles, 
but in PLIST I'd like it to be called IF_NEED_REGEX_H:

${IF_NEED_REGEX_H}include/regex.h

To me, this just _sounds_ good when reading it aloud. :)

> A package requiring, say, regex.h and fnmatch.h would get some code like
> this:
> 
> .if (defined(NEED_REGEX_H) && !empty(NEED_REGEX_H:M[Yy][Ee][Ss])) \
>  || (defined(NEED_FNMATCH_H) && !empty(NEED_FNMATCH_H:M[Yy][Ee][Ss]))
> .include "../../pkgtools/libnbcompat/buildlink3.mk"
> .endif

This looks quite ugly, although it is straight-forward. We are going to
(Continue reading)

Permalink | Reply | 10 comments |
headers
David Griffith | 1 Dec 2005 01:57
Picon
Favicon

pkg/31367


I tried to get this into GNATS, but it didn't seem to go anywhere.
Would someone please help me out?

---------- Forwarded message ----------
Date: Mon, 28 Nov 2005 01:13:31 -0800 (PST)
From: David Griffith <dgriffi <at> cs.csubak.edu>
To: gnats-bugs <at> NetBSD.org
Cc: hauke <at> Espresso.Rhein-Neckar.DE
Subject: pkg/31367

patch-aa should be changed as follows.  However, Wine still won't
work, at least for me.  Anything I do gives this:

$ winemine
wine: failed to initialize: mmap of entire address space failed: Cannot
allocate memory

Anyhow, here's the patch:

--- Make.rules.in.orig  2005-11-28 00:24:48.000000000 -0800
+++ Make.rules.in       2005-11-28 00:25:11.000000000 -0800
 <at>  <at>  -25,7 +25,7  <at>  <at> 
 CC        =  <at> CC <at> 
 CFLAGS    =  <at> CFLAGS <at> 
 CPPFLAGS  =  <at> CPPFLAGS <at> 
-LIBS      =  <at> LIBS <at> 
+LIBS      =  <at> LIBS <at>  -lcompat
 BISON     =  <at> BISON <at> 
 LEX       =  <at> LEX <at>
(Continue reading)

Permalink | Reply | 1 comment |
headers
Roland Illig | 1 Dec 2005 02:32
Picon

Re: pkg/31367

David Griffith wrote:
> I tried to get this into GNATS, but it didn't seem to go anywhere.
> Would someone please help me out?
> 
> ---------- Forwarded message ----------
> Date: Mon, 28 Nov 2005 01:13:31 -0800 (PST)
> From: David Griffith <dgriffi <at> cs.csubak.edu>
> To: gnats-bugs <at> NetBSD.org
> Cc: hauke <at> Espresso.Rhein-Neckar.DE
> Subject: pkg/31367

I think the subject must be "Re: pkg/31367". (Note the "Re:".)

Roland
Permalink | Reply | 0 comments |
headers
Roland Illig | 1 Dec 2005 04:48
Picon

How to compute the checksum of patch files

Hi all,

currently the algorithms used for calculating the checksums of patch 
files differ between pkgtools/pkglint and the rest of pkgsrc. I think 
the pkglint way is "more correct".

others:  ${SED} -e '/\$$NetBSD.*/d' $$patchfile | ${DIGEST} $$alg
pkglint: ${SED} -e '/^\$$NetBSD.*/d' $$patchfile | ${DIGEST} $$alg

For a problematic example where these checksums differ, see devel/error, 
patch-ab:

----snip----
$NetBSD: patch-ab,v 1.1 2005/06/21 21:02:16 jlam Exp $

--- strlcat.c.orig      2005-05-10 16:48:24.000000000 -0400
+++ strlcat.c
 <at>  <at>  -35,6 +35,7  <at>  <at>  __RCSID("$NetBSD: strlcat.c,v 1.16 2003/

  size_t
  strlcat(char *dst, const char *src, size_t siz);
+#undef _DIAGASSERT
  #define _DIAGASSERT(a)

  #ifdef _LIBC
----snip----

In line 5, there is another $NetBSD keyword, so the "others" algorithm 
discards that line from the checksum. I suggest that the two places 
(bsd.pkg.mk and bsd.pkg.patch.mk) are changed to only match at the
(Continue reading)

Permalink | Reply | 12 comments |
headers
Hubert Feyrer | 1 Dec 2005 04:59

Re: How to compute the checksum of patch files

On Thu, 1 Dec 2005, Roland Illig wrote:
> --- strlcat.c.orig      2005-05-10 16:48:24.000000000 -0400
> +++ strlcat.c
>  <at>  <at>  -35,6 +35,7  <at>  <at>  __RCSID("$NetBSD: strlcat.c,v 1.16 2003/

Is adding this standard "diff -u"/pkgdiff behaviour?
I'd rather get rid of that RCS ID there instead of support it.

  - Hubert
Permalink | Reply | 4 comments |
headers
Roland Illig | 1 Dec 2005 05:20
Picon

Re: How to compute the checksum of patch files

Hubert Feyrer wrote:
> On Thu, 1 Dec 2005, Roland Illig wrote:
> 
>> --- strlcat.c.orig      2005-05-10 16:48:24.000000000 -0400
>> +++ strlcat.c
>>  <at>  <at>  -35,6 +35,7  <at>  <at>  __RCSID("$NetBSD: strlcat.c,v 1.16 2003/
> 
> 
> Is adding this standard "diff -u"/pkgdiff behaviour?

Yes, it is.

> I'd rather get rid of that RCS ID there instead of support it.

I generally find it quite useful to see the C function a change is 
happening in. That's what the  <at>  <at>  line is for.

By the way, the $NetBSD could also appear at other places. See 
misc/openoffice/patches/patch-ak, lines 9--10, for an even worse 
example. These lines are excluded from the checksum and could be forged 
without changing the checksum.

Roland
Permalink | Reply | 3 comments |
headers
Hubert Feyrer | 1 Dec 2005 05:36

Re: How to compute the checksum of patch files

On Thu, 1 Dec 2005, Roland Illig wrote:
> By the way, the $NetBSD could also appear at other places. See 
> misc/openoffice/patches/patch-ak, lines 9--10, for an even worse example.

OK
Well, I guess your proposed change (to add a '^') in there makes sense.
Did you do a full "make makepatchsum" on all patches after it to see what 
files really changed? Not that there's any nasty fallout...

  - Hubert
Permalink | Reply | 2 comments |
headers
Roland Illig | 1 Dec 2005 06:04
Picon

Re: How to compute the checksum of patch files

Hubert Feyrer wrote:
> On Thu, 1 Dec 2005, Roland Illig wrote:
> 
>> By the way, the $NetBSD could also appear at other places. See 
>> misc/openoffice/patches/patch-ak, lines 9--10, for an even worse example.
> 
> 
> OK
> Well, I guess your proposed change (to add a '^') in there makes sense.
> Did you do a full "make makepatchsum" on all patches after it to see 
> what files really changed? Not that there's any nasty fallout...

I just used the output of a "pkglint -r /usr/pkgsrc" run, which shows 
all patches where pkglint's and the other's algorithms come to different 
results.

Roland
Permalink | Reply | 1 comment |
headers
Hubert Feyrer | 1 Dec 2005 06:08

Re: How to compute the checksum of patch files

On Thu, 1 Dec 2005, Roland Illig wrote:
>> Did you do a full "make makepatchsum" on all patches after it to see what 
>> files really changed? Not that there's any nasty fallout...
>
> I just used the output of a "pkglint -r /usr/pkgsrc" run, which shows all 
> patches where pkglint's and the other's algorithms come to different results.

So "no". You should, if you intend to commit this.

Out of curiosity, how many files were affected by this - i guess more than 
just the two you mentioned?

  - Hubert
Permalink | Reply | 0 comments |
headers
NetBSD source update | 1 Dec 2005 08:27
Picon

daily pkgsrc CVS update output


Updating pkgsrc tree:
P pkgsrc/archivers/pax/Makefile
P pkgsrc/archivers/pax/files/Makefile.in
P pkgsrc/archivers/pax/files/ar_io.c
P pkgsrc/archivers/pax/files/ar_subs.c
P pkgsrc/archivers/pax/files/buf_subs.c
P pkgsrc/archivers/pax/files/config.h.in
P pkgsrc/archivers/pax/files/configure
P pkgsrc/archivers/pax/files/configure.ac
P pkgsrc/archivers/pax/files/cpio.1
P pkgsrc/archivers/pax/files/cpio.c
P pkgsrc/archivers/pax/files/extern.h
P pkgsrc/archivers/pax/files/file_subs.c
P pkgsrc/archivers/pax/files/ftree.c
P pkgsrc/archivers/pax/files/getoldopt.c
P pkgsrc/archivers/pax/files/options.c
P pkgsrc/archivers/pax/files/pat_rep.c
P pkgsrc/archivers/pax/files/pax.1
P pkgsrc/archivers/pax/files/pax.c
P pkgsrc/archivers/pax/files/pax.h
P pkgsrc/archivers/pax/files/tables.c
P pkgsrc/archivers/pax/files/tables.h
P pkgsrc/archivers/pax/files/tar.1
P pkgsrc/archivers/pax/files/tar.c
P pkgsrc/archivers/pax/files/tty_subs.c
P pkgsrc/audio/gtick/PLIST
P pkgsrc/bootstrap/bootstrap
P pkgsrc/chat/ircservices/Makefile
P pkgsrc/chat/ircservices/PLIST
(Continue reading)

Permalink | Reply | 0 comments |

Gmane