You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@apr.apache.org by Jeff Trawick <tr...@gmail.com> on 2010/02/07 02:47:32 UTC
[VOTE] release apr-1.3.12
See http://apr.apache.org/dev/dist/ for the candidate distribution
files (no Windows source yet -- Bill, is it correct that the one
generated by release.sh is not used?).
The primary changes since apr-1.3.9 are several fixes for the Solaris platform.
+/-1
[ ] Release apr-1.3.12
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Sun, Feb 7, 2010 at 9:19 PM, Gregg L. Smith <li...@glewis.com> wrote:
> Ditto on XP
> All tests pass
> builds fine with httpd 2.2.14 (not that I expected it wouldn't)
> +1
Thanks, all. After a couple of days we have 1 binding vote and
varying degrees of test coverage for the following platforms/builds:
Linux/s390 (SLES9.4 32/64), Linux/PPC (SLES 10.2 32/64), HP/PARISC
(32), HP/IA64 (32/64), AIX61/PPC (32/64), Leopard/x86 (32), Solaris
10U5/x86 (32/64), OpenSolaris 2009.06/x86 (32/64), Windows XP, Windows
7
Customarily these test cycles last 72 hours + delta, so hopefully we
will get a couple of more binding votes in the next day or so. Let's
try to wrap this up by Thursday a.m. US time (unless someone who
intends to test it needs a bit more time).
Re: [VOTE] release apr-1.3.12
Posted by "Gregg L. Smith" <li...@glewis.com>.
Ditto on XP
All tests pass
builds fine with httpd 2.2.14 (not that I expected it wouldn't)
+1
Gregg
Mario Brandt wrote:
> For me it builds still fine like 1.3.11 on Win7.
>
>> (I'd add another step 0 to tag&roll: "make sure it still builds on Windows")
>
>
> Mario
>
Re: [VOTE] release apr-1.3.12
Posted by Mario Brandt <jb...@gmail.com>.
For me it builds still fine like 1.3.11 on Win7.
> (I'd add another step 0 to tag&roll: "make sure it still builds on Windows")
Mario
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Sat, Feb 6, 2010 at 11:04 PM, William A. Rowe Jr.
<wr...@rowe-clan.net> wrote:
> On 2/6/2010 7:47 PM, Jeff Trawick wrote:
>> See http://apr.apache.org/dev/dist/ for the candidate distribution
>> files (no Windows source yet -- Bill, is it correct that the one
>> generated by release.sh is not used?).
>
> Correct, it's used verbatim, with the addition of .mak files which had
> changed early and often in 0.9.
>
> Now that stable releases rarely change source files, my thought is to
> check in win32 command line build files on the .0 tag, and leave them
> in the released branch (touching them only when necessary). This might
> make special 'windows handling' obsolete.
makes sense
(I'd add another step 0 to tag&roll: "make sure it still builds on Windows")
Re: [VOTE] release apr-1.3.12
Posted by "William A. Rowe Jr." <wr...@rowe-clan.net>.
On 2/6/2010 7:47 PM, Jeff Trawick wrote:
> See http://apr.apache.org/dev/dist/ for the candidate distribution
> files (no Windows source yet -- Bill, is it correct that the one
> generated by release.sh is not used?).
Correct, it's used verbatim, with the addition of .mak files which had
changed early and often in 0.9.
Now that stable releases rarely change source files, my thought is to
check in win32 command line build files on the .0 tag, and leave them
in the released branch (touching them only when necessary). This might
make special 'windows handling' obsolete.
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Sat, Feb 6, 2010 at 8:47 PM, Jeff Trawick <tr...@gmail.com> wrote:
> See http://apr.apache.org/dev/dist/ for the candidate distribution
> files (no Windows source yet -- Bill, is it correct that the one
> generated by release.sh is not used?).
>
> The primary changes since apr-1.3.9 are several fixes for the Solaris platform.
>
> +/-1
> [ ] Release apr-1.3.12
+1 for release
looks good on Leopard/x86 (32-bit build), Solaris 10U5/x86 (32-bit and
64-bit builds), OpenSolaris 2009.06/x86 (32-bit and 64-bit builds)
Re: [VOTE] release apr-1.3.12
Posted by Eric Covener <co...@gmail.com>.
On Sat, Feb 6, 2010 at 8:47 PM, Jeff Trawick <tr...@gmail.com> wrote:
> See http://apr.apache.org/dev/dist/ for the candidate distribution
> files (no Windows source yet -- Bill, is it correct that the one
> generated by release.sh is not used?).
>
> The primary changes since apr-1.3.9 are several fixes for the Solaris platform.
>
> +/-1
> [ ] Release apr-1.3.12
>
non-binding +1 for release
clean tests on Linux/s390 (SLES9.4 32/64), Linux/PPC (SLES 10.2
32/64), HP/PARISC (32), HP/IA64 (32/64)
no regression on AIX61/PPC (32/64, testsock failure:
IN6_IS_ADDR_V4MAPPED doesn't match mapped INADDR_ANY)
--
Eric Covener
covener@gmail.com
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Thu, Feb 11, 2010 at 4:06 PM, Jeff Trawick <tr...@gmail.com> wrote:
> On Thu, Feb 11, 2010 at 3:26 PM, William A. Rowe Jr.
> <wr...@rowe-clan.net> wrote:
>> On 2/11/2010 6:10 AM, Jeff Trawick wrote:
>>> On Wed, Feb 10, 2010 at 6:39 PM, Bojan Smojver <bo...@rexursive.com> wrote:
>>>> On Sat, 2010-02-06 at 20:47 -0500, Jeff Trawick wrote:
>>>>> [+1] Release apr-1.3.12
>>>
>>> That's the third binding vote. Barring any other odd reports in the
>>> next 6 or so hours, I'll update the release directory around then.
>>
>> Whoops +1 here; will add win32 source and binaries done this afternoon.
>
> Thanks
>
>> Can I suggest we send out the 1.4.2 email with the security notice, and
>> in _that_ announcement refer to the availability of 1.3.12 for those who do
>> not wish to introduce new functionality? Of course the directory should have
>> both
>
> Does 1.4.2 really need a security notice? (I didn't see anything
> applicable.) Otherwise, "sure."
>
> Do we try to address probable developer expectations that they'd move
> to APR 1.new.x and APR-Util 1.new.x at the same time?
>
> Let's see now...
>
> apr release directory:
>
> create Announcement1.4.*
> (Announcement1.3.* already refers to it as a bug fix release and
> encourages adoption of unnamed "latest APR 1.x")
> HEADER.html: APR 1.4.2 is the latest available version/APR 1.3.12 is
> also available
> README.html: ditto
>
> apr site:
>
> download.* and index.*
>
> I should be free to help in a few hours.
maybe this is a head start (attached)
--
Born in Roswell... married an alien...
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Thu, Feb 11, 2010 at 3:26 PM, William A. Rowe Jr.
<wr...@rowe-clan.net> wrote:
> On 2/11/2010 6:10 AM, Jeff Trawick wrote:
>> On Wed, Feb 10, 2010 at 6:39 PM, Bojan Smojver <bo...@rexursive.com> wrote:
>>> On Sat, 2010-02-06 at 20:47 -0500, Jeff Trawick wrote:
>>>> [+1] Release apr-1.3.12
>>
>> That's the third binding vote. Barring any other odd reports in the
>> next 6 or so hours, I'll update the release directory around then.
>
> Whoops +1 here; will add win32 source and binaries done this afternoon.
Thanks
> Can I suggest we send out the 1.4.2 email with the security notice, and
> in _that_ announcement refer to the availability of 1.3.12 for those who do
> not wish to introduce new functionality? Of course the directory should have
> both
Does 1.4.2 really need a security notice? (I didn't see anything
applicable.) Otherwise, "sure."
Do we try to address probable developer expectations that they'd move
to APR 1.new.x and APR-Util 1.new.x at the same time?
Let's see now...
apr release directory:
create Announcement1.4.*
(Announcement1.3.* already refers to it as a bug fix release and
encourages adoption of unnamed "latest APR 1.x")
HEADER.html: APR 1.4.2 is the latest available version/APR 1.3.12 is
also available
README.html: ditto
apr site:
download.* and index.*
I should be free to help in a few hours.
Re: [VOTE] release apr-1.3.12
Posted by "William A. Rowe Jr." <wr...@rowe-clan.net>.
On 2/11/2010 6:10 AM, Jeff Trawick wrote:
> On Wed, Feb 10, 2010 at 6:39 PM, Bojan Smojver <bo...@rexursive.com> wrote:
>> On Sat, 2010-02-06 at 20:47 -0500, Jeff Trawick wrote:
>>> [+1] Release apr-1.3.12
>
> That's the third binding vote. Barring any other odd reports in the
> next 6 or so hours, I'll update the release directory around then.
Whoops +1 here; will add win32 source and binaries done this afternoon.
Can I suggest we send out the 1.4.2 email with the security notice, and
in _that_ announcement refer to the availability of 1.3.12 for those who do
not wish to introduce new functionality? Of course the directory should have
both
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Wed, Feb 10, 2010 at 6:39 PM, Bojan Smojver <bo...@rexursive.com> wrote:
> On Sat, 2010-02-06 at 20:47 -0500, Jeff Trawick wrote:
>> [+1] Release apr-1.3.12
That's the third binding vote. Barring any other odd reports in the
next 6 or so hours, I'll update the release directory around then.
Thanks again, everyone!
Re: [VOTE] release apr-1.3.12
Posted by Bojan Smojver <bo...@rexursive.com>.
On Sat, 2010-02-06 at 20:47 -0500, Jeff Trawick wrote:
> [+1] Release apr-1.3.12
Signatures look good. Checksums looks good.
All tests pass on Fedora 12, i686.
--
Bojan
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Wed, Feb 10, 2010 at 5:37 PM, Gregory Szorc <gr...@gmail.com> wrote:
> I'm not sure if this is the best place to post this, but when running `make
> check` against the 1.3.12 dist on Solaris x86, the test hangs in testpoll.
> Attaching a debugger reveals the following backtrace:
>
> #0 0xfedc9a45 in _portfs () from /lib/libc.so.1
> #1 0xfed55c29 in port_getn () from /lib/libc.so.1
> #2 0xfef898a6 in apr_pollcb_poll (pollcb=0x8163238, timeout=-4160455976,
> func=0x806626d <trigger_pollcb_cb>,
> baton=0x8047b04) at poll/unix/port.c:490
> #3 0x08066408 in trigger_pollcb (tc=0x8047b3c, data=0x0) at testpoll.c:617
> #4 0x080557c1 in abts_run_test (ts=0x80f5b20, f=0x80662f1 <trigger_pollcb>,
> value=0x0) at abts.c:168
> #5 0x08066904 in testpoll (suite=0x80f5b20) at testpoll.c:702
> #6 0x08056171 in main (argc=2, argv=0x8047bc4) at abts.c:424
>
> uname is "SunOS 5.10 Generic_142901-03 i86pc i386 i86pc." /etc/release
> reports as "Solaris 10 10/09 s10x_u8wos_08a X86."
I forgot that I had that level of Solaris 10 too, but I guess you have
some kernel patch applied on top?
-bash-3.00$ uname -a
SunOS unknown 5.10 Generic_141445-09 i86pc i386 i86pc
-bash-3.00$ head -1 /etc/release
Solaris 10 10/09 s10x_u8wos_08a X86
-bash-3.00$ cd test
-bash-3.00$ ./testall -v testpoll
testpoll : SUCCESS
All tests passed.
(succeeds repeatedly with" while ./testall -v testpoll; do :; done")
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
(oops, sent to Gregory directly)
On Wed, Feb 10, 2010 at 6:25 PM, Jeff Trawick <tr...@gmail.com> wrote:
> On Wed, Feb 10, 2010 at 5:37 PM, Gregory Szorc <gr...@gmail.com> wrote:
>> I'm not sure if this is the best place to post this, but when running `make
>> check` against the 1.3.12 dist on Solaris x86, the test hangs in testpoll.
>
> works for me on S10U5 (a.k.a. 05/08)
>
>> Attaching a debugger reveals the following backtrace:
>>
>> #0 0xfedc9a45 in _portfs () from /lib/libc.so.1
>> #1 0xfed55c29 in port_getn () from /lib/libc.so.1
>> #2 0xfef898a6 in apr_pollcb_poll (pollcb=0x8163238, timeout=-4160455976,
>
> weird value displayed for timeout (equiv to 0x80000000, right?);
> should be -1 (-1LL)
>
> what does "pstack PID" display when it is hung?
>
>> func=0x806626d <trigger_pollcb_cb>,
>> baton=0x8047b04) at poll/unix/port.c:490
>> #3 0x08066408 in trigger_pollcb (tc=0x8047b3c, data=0x0) at testpoll.c:617
>> #4 0x080557c1 in abts_run_test (ts=0x80f5b20, f=0x80662f1 <trigger_pollcb>,
>> value=0x0) at abts.c:168
>> #5 0x08066904 in testpoll (suite=0x80f5b20) at testpoll.c:702
>> #6 0x08056171 in main (argc=2, argv=0x8047bc4) at abts.c:424
>>
>> uname is "SunOS 5.10 Generic_142901-03 i86pc i386 i86pc." /etc/release
>> reports as "Solaris 10 10/09 s10x_u8wos_08a X86."
>>
>> If someone can point me to the appropriate bug, I'll post whatever
>> information I can.
>
> I don't think a bug is open for this symptom.
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Thu, Feb 11, 2010 at 1:29 AM, Gregory Szorc <gr...@gmail.com> wrote:
>> Sure; it might be quicker to use LD_LIBRARY_PATH to run the 1.3.12
>> testall testpoll against apr 1.3.9's libapr before sorting through
>> individual commits.
>
> The instant I read this, I realized my shell likely had LD_LIBRARY_PATH set.
> Sure enough, it was. And, it was pointing to a path that had a 1.3.8 APR
> shared library. After unsetting LD_LIBRARY_PATH, `testall` found the proper
> 1.3.12 library and passed the aforementioned poll tests.
>
> I suppose testall could be compiled to reference libapr-1.so.x.y.z instead
> of libapr-1.so to minimize the possibility of this. But, a bad environment
> is a bad environment.
>
> Sorry for the fire drill. I'll try not to be so careless next time.
OTOH, apr releases tend to be a bit boring.
Re: [VOTE] release apr-1.3.12
Posted by Gregory Szorc <gr...@gmail.com>.
> Sure; it might be quicker to use LD_LIBRARY_PATH to run the 1.3.12
> testall testpoll against apr 1.3.9's libapr before sorting through
> individual commits.
The instant I read this, I realized my shell likely had LD_LIBRARY_PATH set.
Sure enough, it was. And, it was pointing to a path that had a 1.3.8 APR
shared library. After unsetting LD_LIBRARY_PATH, `testall` found the proper
1.3.12 library and passed the aforementioned poll tests.
I suppose testall could be compiled to reference libapr-1.so.x.y.z instead
of libapr-1.so to minimize the possibility of this. But, a bad environment
is a bad environment.
Sorry for the fire drill. I'll try not to be so careless next time.
Greg
Re: [VOTE] release apr-1.3.12
Posted by Jeff Trawick <tr...@gmail.com>.
On Wed, Feb 10, 2010 at 9:50 PM, Gregory Szorc <gr...@gmail.com> wrote:
>> what does "pstack PID" display when it is hung?
>
> 3305: ./testall testpoll
> fedc9a45 portfs (6, 37, 8110b98, 32, 32, 0)
> fef898a6 apr_pollcb_poll (8110b88, ffffffff, ffffffff, 806626d, 8047b04,
> 8047b3c) + 82
> 08066408 trigger_pollcb (8047b3c, 0, 80772eb, 0, 80f55d8, 80f55d8) + 117
> 080557c1 abts_run_test (80f5b20, 80662f1, 0, 0, 807384d, 8088eac) + 56
> 08066904 testpoll (80f5b20, 4, fefcab34, 8047ba4, 16, 807512a) + 1fa
> 08056171 main (8055170, 2, 8047bbc) + 20f
> 08055170 _start (2, 8047ca8, 8047cb2, 0, 8047cbb, 8047cf3) + 80
Hmmm... If I comment out the send_msg() call so that there's no data
available yet, my backtrace looks like
25072: ./testall testpoll
fee04157 portfs (6, 37, 81155a0, 32, 1, 0)
fef87218 call_port_getn (37, 81155a0, 32, 8047180, ffffffff, ffffffff) + c8
fef881a4 apr_pollcb_poll (8115590, ffffffff, ffffffff, 8066c40,
80471c0, 0) + 54
08066dde trigger_pollcb (80471fc, 0) + fe
08056331 abts_run_test (80fa528, 8066ce0, 0) + 71
08067274 testpoll (80fa528, 0) + 184
08056ee3 main (2, 8047284, 8047290, 8055c8f) + 213
08055ced _start (2, 80473b8, 80473c2, 0, 80473cb, 804743d) + 7d
Interestingly, the fifth parm to portfs() in your backtrace is 32 ==
nalloc, and the fifth parm to portfs() in mine is 1. Meanwhile,
there's a bug fix in 1.3.12 to fix a hang in apr_pollcb_poll() on
Solaris by passing 1 instead of nalloc for the number of events to
wait for.
I can't imagine how you wouldn't have the fix or would be running the
wrong libapr, but can you check with pldd which libapr is getting
loaded just in case?
(I'm guessing the absence of call_port_getn() in the backtrace is due
to gcc inlining, though apr <= 1.3.8 doesn't have that function.)
> I have access to other Solaris releases. I can always try to compile and
> test on them. I may also start going through the SVN commits and isolating
> the failure to a specific revision. Of course, this could all be related to
> my toolchain - I'm using GNU for everything but the linker. Still, a bug is
> a bug.
Sure; it might be quicker to use LD_LIBRARY_PATH to run the 1.3.12
testall testpoll against apr 1.3.9's libapr before sorting through
individual commits.
(so weird)
Re: [VOTE] release apr-1.3.12
Posted by Gregory Szorc <gr...@gmail.com>.
I'm not sure if this is the best place to post this, but when running `make
check` against the 1.3.12 dist on Solaris x86, the test hangs in testpoll.
Attaching a debugger reveals the following backtrace:
#0 0xfedc9a45 in _portfs () from /lib/libc.so.1
#1 0xfed55c29 in port_getn () from /lib/libc.so.1
#2 0xfef898a6 in apr_pollcb_poll (pollcb=0x8163238, timeout=-4160455976,
func=0x806626d <trigger_pollcb_cb>,
baton=0x8047b04) at poll/unix/port.c:490
#3 0x08066408 in trigger_pollcb (tc=0x8047b3c, data=0x0) at testpoll.c:617
#4 0x080557c1 in abts_run_test (ts=0x80f5b20, f=0x80662f1 <trigger_pollcb>,
value=0x0) at abts.c:168
#5 0x08066904 in testpoll (suite=0x80f5b20) at testpoll.c:702
#6 0x08056171 in main (argc=2, argv=0x8047bc4) at abts.c:424
uname is "SunOS 5.10 Generic_142901-03 i86pc i386 i86pc." /etc/release
reports as "Solaris 10 10/09 s10x_u8wos_08a X86."
If someone can point me to the appropriate bug, I'll post whatever
information I can.
Greg
On Wed, Feb 10, 2010 at 1:46 AM, Joe Orton <jo...@redhat.com> wrote:
> On Sat, Feb 06, 2010 at 08:47:32PM -0500, Jeff Trawick wrote:
> > See http://apr.apache.org/dev/dist/ for the candidate distribution
> > files (no Windows source yet -- Bill, is it correct that the one
> > generated by release.sh is not used?).
> >
> > The primary changes since apr-1.3.9 are several fixes for the Solaris
> platform.
> >
> > +/-1
> > [+1] Release apr-1.3.12
>
> Looks good here - builds and passes tests on Fedora 12/x86_64.
>
> - Diff vs 1.3.9 looks fine
> - CHANGES looks fine
> - Signatures fine
>
> Thanks for RMing!
>
> Regards, Joe
>
Re: [VOTE] release apr-1.3.12
Posted by Joe Orton <jo...@redhat.com>.
On Sat, Feb 06, 2010 at 08:47:32PM -0500, Jeff Trawick wrote:
> See http://apr.apache.org/dev/dist/ for the candidate distribution
> files (no Windows source yet -- Bill, is it correct that the one
> generated by release.sh is not used?).
>
> The primary changes since apr-1.3.9 are several fixes for the Solaris platform.
>
> +/-1
> [+1] Release apr-1.3.12
Looks good here - builds and passes tests on Fedora 12/x86_64.
- Diff vs 1.3.9 looks fine
- CHANGES looks fine
- Signatures fine
Thanks for RMing!
Regards, Joe