On 11/12/11 06:36 AM, Norm Jacobs wrote:
As a general comment, a little extra whitespace (at the begining of
continued lines, before +=, ...) might make it slightly more readable.
components/ksh93/Makefile
Lines 27-28
You aren't using these, but you might put a comment about them
above.
Done.
Lines 51-56
I would rather you just turn of '-B direct' with LD_B_DIRECT=
Also done.
Line 61
remove this and use CXXFLAGS in attpackagemake.mk:49,55
instead. We already set the flags we want including -m$(BITS)
I was having serious issues with that - I'll try again now I have a
better understanding.
Lines 62-63
You aren't using these and should not need them.
I thought so, but I think this is related to the CXXFLAGS above.
make-rules/attpackagemake.mk
Lines 47-49:
LD_OPTIONS is set in shared-macros.mk
CCFLAGS is equivalent to CXXFLAGS in shared-macros.mk, just us it.
That one I'd missed :-)
Line 53:
Use += instead of =
Also that one...
Line 54:
This is already added in shared-macros.mk
Line55-56:
use CCFLAGS=$(CXXFLAGS)
and CCLDFLAGS=$(CXXFLAGS)
That should sort the majority of my issues - AT&T's nmake uses a
*variety* of defines.....
Lines 58-59:
This is already added in shared-macros.mk
Yep.
Lines 62-63:
You might look at shared-macros.mk again.
make-rules/attprep.mk
I was hoping that this might get rolled into prep.mk. So that we
only had one makefile fragment for archive download, unpack, and
patch regardless of whether it was single or multi-archive.
Hi there - I originally tried to keep the two to behave similarly,
but then ran up against the issue of being able to patch each archive
individually, specifying which patches apply to which archive: this
actually breaks the current patching philosophy of searching the
directory and applying - I may be able to make the two ways cooperate,
but figured that for simplicity's sake, keep it separate and integrate
the two after the initial ksh93 integration.
I assume that the license, manifest, and patches pretty much match
what's in ON, so I haven't looked at them closely, though I am
wondering if there are any docs that move with this?
As far as we have worked out, the idea is to be identical to the ON
functionality and packaging - this should result in no Doc changes
(though I believe it is usual for the userland components to deliver
their own docs?).
-Norm
On 11/11/11 03:17 AM, Edwin Beasant wrote:
Hi there, I'd like to request a code review for the introduction of
Ksh93 into userland from ON
http://monaco.us.oracle.com/detail.jsf?cr=7106955
Webrev is here:
http://odin.uk.oracle.com/~eb243102/webrevs/ksh_userland_v5/
Description:
This change brings in additions to the make-rules suite:
1. Addition of rules to allow building of AT&T style 'package'
builds.
2. Addition of rules to allow download, unpack and patch to be
separately controllable for multiple tarballs that may be required to
form a complete source tree (a requirement of the AT&T package style).
It also comprises ported patches for the following bugs (fixed in the
ON version):
#RM <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C7032821%2C7036535%2C7065478%2C7065900%2C7071431&csort=responsible_manager> RE <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C7032821%2C7036535%2C7065478%2C7065900%2C7071431&csort=responsible_engineer> CAT/SUBCAT <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C7032821%2C7036535%2C7065478%2C7065900%2C7071431&csort=cat_subcat> P <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C7032821%2C7036535%2C7065478%2C7065900%2C7071431&csort=priority> ID <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C7032821%2C7036535%2C7065478%2C7065900%2C7071431&cso%0Art=cr_number> STA <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C703%0A2821%2C7036535%2C7065478%2C7065900%2C7071431&csort=status> AGE <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C7032821%2C7036535%2C7065478%2C7065900%2C7071431&csort=age> SYNOPSIS <http://monaco.us.oracle.com/list.jsf?cr=6919590%2C6964621%2C7025778%2C7026179%2C7032821%2C7036535%2C7065478%2C7065900%2C7071431&csort=synopsis>
1 <http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=7065900%0A%0Aset+OrigCR=7065900%0A%0Ado+query/raw> bonnie.corwin <mailto:bonnie.cor...@oracle.com> alan.burlison <mailto:alan.burli...@oracle.com> shell/korn93 37065900 <http://monaco.us.oracle.com/detail.jsf?cr=7065900> fde 4M ksh93 builtins should be pruned and uses of them removed
2
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=7071431%0A%0Aset+OrigCR=7071431%0A%0Ado+query/raw>
joe.g <mailto:jo...@oracle.com> alan.burlison
<mailto:alan.burli...@oracle.com> shell/korn93 17071431
<http://monaco.us.oracle.com/detail.jsf?cr=7071431> fde 3M ksh93 ulimit lost support for
many options
3
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=7026179%0A%0Aset+OrigCR=7026179%0A%0Ado+query/raw>
joe.g <mailto:jo...@oracle.com> amrita.sadhukha
<mailto:amrita.sadhuk...@oracle.com> utility/file 37026179
<http://monaco.us.oracle.com/detail.jsf?cr=7026179> fde 8M wc /usr/pub/utf-8 segfaults
4
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=6964621%0A%0Aset+OrigCR=6964621%0A%0Ado+query/raw>
joe.g <mailto:jo...@oracle.com> nobutomo.nakano
<mailto:nobutomo.nak...@oracle.com> shell/korn93 26964621
<http://monaco.us.oracle.com/detail.jsf?cr=6964621> fde 16M ksh93 core found on the system
5
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=6919590%0A%0Aset+OrigCR=6919590%0A%0Ado+query/raw>
joe.g <mailto:jo...@oracle.com> ralph.turner
<mailto:ralph.tur...@oracle.com> shell/korn93 26919590
<http://monaco.us.oracle.com/detail.jsf?cr=6919590>-M fde 21M Problem with shell/korn93
6
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=7065478%0A%0Aset+OrigCR=7065478%0A%0Ado+query/raw>
joe.g <mailto:jo...@oracle.com> ralph.turner
<mailto:ralph.tur...@oracle.com> shell/korn93 27065478
<http://monaco.us.oracle.com/detail.jsf?cr=7065478> fde 4M ksh93 coredump
sun_solaris_cr_6805794_character_to_wchar_not_working(ja_JP.eucJP)
7
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=7025778%0A%0Aset+OrigCR=7025778%0A%0Ado+query/raw>
venkateshwara.t <mailto:venkateshwara....@oracle.com> nobutomo.nakano
<mailto:nobutomo.nak...@oracle.com> utility/file 47025778
<http://monaco.us.oracle.com/detail.jsf?cr=7025778>-M fde 8M cmp(1) against two ZFS directories
or a ZFS Directory and 0byte file returns exit 0 (i.e. Match)
8
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=7036535%0A%0Aset+OrigCR=7036535%0A%0Ado+query/raw>
venkateshwara.t <mailto:venkateshwara....@oracle.com> ralph.turner
<mailto:ralph.tur...@oracle.com> shell/korn93 17036535
<http://monaco.us.oracle.com/detail.jsf?cr=7036535> fde 7M ksh93 update causes segv in
path_addpath
9
<http://hestia.sfbay.sun.com/cgi-bin/expert?Go=2;cmds=set+CR=7032821%0A%0Aset+OrigCR=7032821%0A%0Ado+query/raw>
venkateshwara.t <mailto:venkateshwara....@oracle.com> venkateshwara.t
<mailto:venkateshwara....@oracle.com> shell/korn93 17032821
<http://monaco.us.oracle.com/detail.jsf?cr=7032821> fde 7M Changes in the Korn shell in
snv_163 cause segmentation faults
Additionally the following code change was not attributable to a
single CR in ON:
CRXXX_Error_Catalog.patch
The following patches fix issues with the build of ksh93 and
supporting AT&T tools:
build_cflags.patch
multi_lang_arith.patch
path_utmp.patch
The following is currently unused, but is required if the ksh93
package is to be minimally built (i.e. without remainder of the
ast-base package).
package-pax-suid.patch
The following patch was introduced to provide the current Solaris
'/usr/bin/alias' functionality:
solaris_alias.patch
The AT&T ksh93 is contained in an AT&T package 'ast-base' which in
turn requires 'INIT' to function/compile - both of these tarballs
have been pre-seeded to the userland repository.
Currently, the patch to remove the 'sleep' builtin is unused, as it
causes issues with ksh93 testsuite, which has also been integrated
into the userland system: a bug is to be filed to cover the work to
fix the ksh93 test suite so that it functions with the Solaris
version of sleep.
Many Thanks,
Edwin Beasant.
_______________________________________________
userland-discuss mailing list
userland-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/userland-discuss
_______________________________________________
userland-discuss mailing list
userland-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/userland-discuss
