Re: [PATCH 00/21] add ia64 to ports & update to latest tree

[Mike Frysinger <vapier at gentoo dot org>]

On Wednesday 25 April 2012 00:12:42 Carlos O'Donell wrote:
> On Mon, Apr 23, 2012 at 9:57 PM, Mike Frysinger <vapier@gentoo.org> wrote:
> > On Monday 23 April 2012 21:43:39 Carlos O'Donell wrote:
> >> On Mon, Apr 23, 2012 at 8:03 PM, Mike Frysinger wrote:
> >> > On Saturday 21 April 2012 00:35:21 Mike Frysinger wrote:
> >> >> The `make check` results on my ia64 system seems to be just as good
> >> >> as the glibc-2.15 release.  In other words, we get a few failures,
> >> >> but no regressions.  In fact, some tests now pass that didn't before
> >> >> :).
> >> >> 
> >> >> I also compared the exported symbol list:
> >> >>       for l in `find -type l` ; do
> >> >>               readelf -sW /lib/$l | awk '$7 != "UND" { print $NF }' |
> >> >> sort > old readelf -sW $l | awk '$7 != "UND" { print $NF }' | sort >
> >> >> new diff -u old new
> >> >>       done
> >> >> Ignoring the new GLIBC_2.16 symbols, the lists were the same.
> >> >> 
> >> >> Mike Frysinger (21):
> >> >>   ia64: move from main tree
> >> >>   ia64: strip trailing whitespace
> >> >>   ia64: update FSF license address to URL
> >> >>   ia64: fix license text to "Lesser" not "Library"
> >> >>   ia64: update include paths
> >> >>   ia64: add split-out settings into dedicated files
> >> >>   ia64: merge sysdeps/ia64/elf/ into sysdeps/ia64/
> >> >>   ia64: initfini.c -> crt{i,n}.S split
> >> >>   ia64: convert <stdio-common/_itoa.h> to <_itoa.h>
> >> >>   ia64: convert "math{,_private}.h" to <math{,_private}.h>
> >> >>   ia64: convert __const to const
> >> >>   ia64: math: stub out new e_rem_pio2l func
> >> >>   ia64: drop __STDC__ handlings
> >> >>   ia64: math: add __scalbn* aliases
> >> >>   ia64: add MAP_{STACK,HUGETLB} to bits/mman.h
> >> >>   ia64: pthread_attr_t type mangling
> >> >>   ia64: ucontext.h: drop __cplusplus
> >> >>   ia64: splitting of 16bit funcs out of byteswap.h and into
> >> >>     byteswap-16.h
> >> >>   ia64: add new GET_ADDR_PARAM define
> >> >>   ia64: define __ASSUME_ACCEPT4
> >> >>   ia64: math: rename s_ilogbl.S to e_ilogbl.S
> >> > 
> >> > if there's no feedback by the end of the week, i'l go ahead and merge
> >> > the rest
> >> 
> >> Please create and update ChangeLog.ia64 with your GNU ChangeLog entries
> >> :-)
> > 
> > that's what i've been doing with the commits i've pushed so far ... was
> > there something else you meant ?
> 
> No, I must have not looked closely enough at the commit emails.

perhaps the confusion arises from these git patches containing the changelog 
details in the git commit message instead of actually patching the ChangeLog 
file.  with other GNU projects i've worked with, including a diff against the 
ChangeLog is undesirable, so i've gotten in the habit of including in the 
commit message and then when actually making the commit, moving it to the 
right file.  this has also certainly been easier to maintain due to the 
constant conflicts that come from trying to rebase on a file that changes as 
frequently as these do.
-mike

Attachment: signature.asc
Description: This is a digitally signed message part.