136163 - get mozilla branch building on ipf

Status: RESOLVED FIXED

(SeaMonkey :: Build Config, defect)

Description

[Jim Dunn]

This is a place holder for getting mozilla 6.2.2 building on 
IPF (hp-ux 11.20 - ia64)

Comment 1

[Jim Dunn]

updating

Comment 2

[Jim Dunn]

Attachment: 1st set of patches to get mozilla 622 working on ipf

This patch contains most of the build changes required
to get MOZILLA_6_2_2_BRANCH working on ipf (hpux 11.20 on ia64)
using aCC x.05.32.09.

does not include all of the nspr files

Comment 3

[Jim Dunn]

Comment on attachment 78188 [details] [diff] [review]
1st set of patches to get mozilla 622 working on ipf

patches checked into MOZILLA_6_2_2_BRANCH

Comment 4

[timeless]

did you really check in bare printfs? ah, yes you did.
http://bonsai.mozilla.org/cvsblame.cgi?file=mozilla/security/nss/lib/util/unix_rand.c&rev=NETSCAPE_6_2_2_BRANCH&mark=#1
does that make sense?

Comment 5

[Jim Dunn]

Attachment: additional changes

Adding the rules.mk change to only export the required symbols
from Components.  This gives us a 10% reduction in binary size
and should improve loading performance.

Also change from using _pr_CopyLowBits to CopyLowBits.	_pr
is in nspr and "shouldn't" be exported whereas we should be
using the call inside of sec utility.

Comment 6

[Jim Dunn]

updating

Comment 7

[Martin Lawyer]

Attachment: Latest complete patch set

As of 8/16, this cumulative patch is sufficient to make the OEM branch build on
IPF.

Comment 8

[Martin Lawyer]

Comment on attachment 95934 [details] [diff] [review]
Latest complete patch set

Need to keep some of these patches separate. Ignore this.

Comment 9

[Martin Lawyer]

Attachment: OEM branch patch

Comment 10

[Martin Lawyer]

Changes checked into OEM branch.

Comment 11

[Jim Dunn]

Attachment: latest diff

The other parts of the original patch have been picked
up by other bugs.  This one patch is the major one needed
for this bug now.

Comment 12

[Jim Dunn]

adding cls&jband because they can r=/sr= the changes to
xpcom/reflect/xptcall/src/unix/Makefile.in

I think I can just get away with cls's, but added jband just
to keep him in the loop.

cls, what do you think?

Comment 13

[hacker formerly known as [e-mail]]

Comment on attachment 100247 [details] [diff] [review]
latest diff

r=cls

Comment 14

[Jim Dunn]

fix checked into trunk