[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <000001cb6776$97e573f0$c7b05bd0$@net>
Date: Sat, 9 Oct 2010 01:55:36 -0400
From: "Robert Harris" <rs904c@...scape.net>
To: <john-users@...ts.openwall.com>
Subject: RE: Patch for Makefile, concerning Solaris x86
Alex and John-Users,
We does the tgtsnarf program do, anyway? Why is it only built in Solaris?
So that there is one good Makefile for Solaris and all other builds...
I purpose we either do this for each Solaris build item (What is a better
name for this?), Option 1)
(add the text "-lnsl -lsocket" to each Solaris build item's LDFLAGS)
solaris-x86-sse2-gcc:
$(LN) x86-sse.h arch.h
$(MAKE) $(PROJ) \
JOHN_OBJS="$(JOHN_OBJS) c3_fmt.o solaris-x86.o x86-sse.o
sha1-mmx.o md5-mmx.o" \
CFLAGS="$(CFLAGS) -DHAVE_CRYPT" \
LDFLAGS="$(LDFLAGS) -lrt -lnsl -lsocket"
or this for each Solaris item Option 2)
(create a new global variable for Solaris only, and use that in the LDFLAGS)
(Looks like you have a MARKOV LDFLAGS, why not a new one for Solaris?)
LDFLAGS_SOLARIS = -lnsl -lsocket (New Global variable)
solaris-x86-sse2-gcc:
$(LN) x86-sse.h arch.h
$(MAKE) $(PROJ) \
JOHN_OBJS="$(JOHN_OBJS) c3_fmt.o solaris-x86.o x86-sse.o
sha1-mmx.o md5-mmx.o" \
CFLAGS="$(CFLAGS) -DHAVE_CRYPT" \
LDFLAGS="$(LDFLAGS_SOLARIS) -lrt"
Which do you prefer? Looks like either one works.
-Robert Harris
-----Original Message-----
From: Solar Designer [mailto:solar@...nwall.com]
Sent: Friday, October 08, 2010 5:56 PM
To: john-users@...ts.openwall.com
Subject: Re: [john-users] Patch for Makefile, concerning Solaris x86
Robert,
On Fri, Oct 08, 2010 at 05:42:58PM -0400, Robert Harris wrote:
> I was thinking about doing it in the global LDFLAGS, but that would affect
> all builds. Wouldn't that cause a problem? (I haven't checked yet, if it
> would).
Of course, it would break builds on most non-Solaris systems.
I'm sorry, I've used poor wording below. By "global" LDFLAGS, I meant
non-program-specific ones - that is, overrides for the LDFLAGS value
that you specify in solaris-* targets, as opposed to overrides of some
new make variable that would be specific to the tgtsnarf program (such a
variable doesn't exist currently). I argued that using the existing
non-program-specific LDFLAGS was sort of OK this time because the jumbo
patch was already doing it for other libraries that would otherwise need
to be program-specific. If we're to improve upon this, we'd need to do
it for more than just tgtsnarf anyway.
> From: Solar Designer [mailto:solar@...nwall.com]
...
> All solaris-* targets need it (this means solaris-sparc* and *-cc ones
> as well), albeit only for the tgtsnarf program in the jumbo patch. It
> is easier to add the options to the global LDFLAGS rather than to try
> to do it for tgtsnarf only. In fact, the jumbo patch is already dirty
> in this aspect anyway - it similarly unnecessarily links some programs
> against OpenSSL. So we can continue this "tradition" for now...
Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
