You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@httpd.apache.org by Stefan Eissing <st...@greenbytes.de> on 2017/04/27 14:15:51 UTC
backport proposals
Take this as an observation about proposals in general, nothing wrong with this one in particular:
*) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging
trunk patch: http://svn.apache.org/r1784203
http://svn.apache.org/r1784205
http://svn.apache.org/r1784227
http://svn.apache.org/r1784228
http://svn.apache.org/r1784275
http://svn.apache.org/r1785871
http://svn.apache.org/r1786009
http://svn.apache.org/r1789387
2.4.x patch: trunk works *after r1779573 above* (modulo CHANGES)
ie: http://home.apache.org/~ylavic/patches/httpd2.4-hcheck-after-r1779573.patch
FULL hcheck patch: http://home.apache.org/~jim/patches/httpd2.4-hcheck.patch
http://svn.apache.org/r1789387
(includes all hcheck related patches, including showstopper)
+1: jim, ylavic
So, how to check this? I tried the FULL hcheck patch. It does not work, ~50% of hunks fail.
Ok, revert. Then I just tried this:
svn merge -c 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387 ^/httpd/httpd/trunk .
and all is well! So, the proposal could have been written as:
*) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging
svn merge -c 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387 ^/httpd/httpd/trunk .
+1: jim, ylavic
Wouldn't that be easier? I mean, sometimes trunk and backport may differ a lot. But most commonly, only CHANGES and message-tags need to be ignored. I myself would prefer just to copy&exec a one liner.
just saying (while the active hcheck runs in my proxy tests flawlessly),
-Stefan
Re: backport proposals
Posted by Daniel Ruggeri <dr...@primary.net>.
Which reminds me... How about that cool proxy protocol patch? Anyone want to give it a whirl? ;-)
--
Daniel Ruggeri
-------- Original Message --------
From: Jim Jagielski <ji...@jaguNET.com>
Sent: April 27, 2017 9:55:39 AM CDT
To: dev@httpd.apache.org
Subject: Re: backport proposals
Yeah... sometimes it is easier to create a "combined" patchfile
which shows the collected changes, for ease of reviewing, but, as
you say, it can get out of sync as other patches are backported,
esp if the proposed backport is in STATUS for a semi-extended
period of time.
> On Apr 27, 2017, at 10:47 AM, Eric Covener <co...@gmail.com> wrote:
>
> On Thu, Apr 27, 2017 at 10:15 AM, Stefan Eissing
> <st...@greenbytes.de> wrote:
>> Take this as an observation about proposals in general, nothing wrong with this one in particular:
>>
>> *) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging
>> trunk patch: http://svn.apache.org/r1784203
>> http://svn.apache.org/r1784205
>> http://svn.apache.org/r1784227
>> http://svn.apache.org/r1784228
>> http://svn.apache.org/r1784275
>> http://svn.apache.org/r1785871
>> http://svn.apache.org/r1786009
>> http://svn.apache.org/r1789387
>> 2.4.x patch: trunk works *after r1779573 above* (modulo CHANGES)
>> ie: http://home.apache.org/~ylavic/patches/httpd2.4-hcheck-after-r1779573.patch
>> FULL hcheck patch: http://home.apache.org/~jim/patches/httpd2.4-hcheck.patch
>> http://svn.apache.org/r1789387
>> (includes all hcheck related patches, including showstopper)
>> +1: jim, ylavic
>>
>> So, how to check this? I tried the FULL hcheck patch. It does not work, ~50% of hunks fail.
>>
>> Ok, revert. Then I just tried this:
>> svn merge -c 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387 ^/httpd/httpd/trunk .
>
> Looks much easier to me. Another option is what kotkov recently used
>
> Not 100% clear to me in this case that the collected patch was really
> just meant to be the same revs + merge conflicts due to the wording.
Re: backport proposals
Posted by Jim Jagielski <ji...@jaguNET.com>.
Yeah... sometimes it is easier to create a "combined" patchfile
which shows the collected changes, for ease of reviewing, but, as
you say, it can get out of sync as other patches are backported,
esp if the proposed backport is in STATUS for a semi-extended
period of time.
> On Apr 27, 2017, at 10:47 AM, Eric Covener <co...@gmail.com> wrote:
>
> On Thu, Apr 27, 2017 at 10:15 AM, Stefan Eissing
> <st...@greenbytes.de> wrote:
>> Take this as an observation about proposals in general, nothing wrong with this one in particular:
>>
>> *) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging
>> trunk patch: http://svn.apache.org/r1784203
>> http://svn.apache.org/r1784205
>> http://svn.apache.org/r1784227
>> http://svn.apache.org/r1784228
>> http://svn.apache.org/r1784275
>> http://svn.apache.org/r1785871
>> http://svn.apache.org/r1786009
>> http://svn.apache.org/r1789387
>> 2.4.x patch: trunk works *after r1779573 above* (modulo CHANGES)
>> ie: http://home.apache.org/~ylavic/patches/httpd2.4-hcheck-after-r1779573.patch
>> FULL hcheck patch: http://home.apache.org/~jim/patches/httpd2.4-hcheck.patch
>> http://svn.apache.org/r1789387
>> (includes all hcheck related patches, including showstopper)
>> +1: jim, ylavic
>>
>> So, how to check this? I tried the FULL hcheck patch. It does not work, ~50% of hunks fail.
>>
>> Ok, revert. Then I just tried this:
>> svn merge -c 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387 ^/httpd/httpd/trunk .
>
> Looks much easier to me. Another option is what kotkov recently used
>
> Not 100% clear to me in this case that the collected patch was really
> just meant to be the same revs + merge conflicts due to the wording.
Re: backport proposals
Posted by Eric Covener <co...@gmail.com>.
On Thu, Apr 27, 2017 at 10:15 AM, Stefan Eissing
<st...@greenbytes.de> wrote:
> Take this as an observation about proposals in general, nothing wrong with this one in particular:
>
> *) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging
> trunk patch: http://svn.apache.org/r1784203
> http://svn.apache.org/r1784205
> http://svn.apache.org/r1784227
> http://svn.apache.org/r1784228
> http://svn.apache.org/r1784275
> http://svn.apache.org/r1785871
> http://svn.apache.org/r1786009
> http://svn.apache.org/r1789387
> 2.4.x patch: trunk works *after r1779573 above* (modulo CHANGES)
> ie: http://home.apache.org/~ylavic/patches/httpd2.4-hcheck-after-r1779573.patch
> FULL hcheck patch: http://home.apache.org/~jim/patches/httpd2.4-hcheck.patch
> http://svn.apache.org/r1789387
> (includes all hcheck related patches, including showstopper)
> +1: jim, ylavic
>
> So, how to check this? I tried the FULL hcheck patch. It does not work, ~50% of hunks fail.
>
> Ok, revert. Then I just tried this:
> svn merge -c 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387 ^/httpd/httpd/trunk .
Looks much easier to me. Another option is what kotkov recently used
Not 100% clear to me in this case that the collected patch was really
just meant to be the same revs + merge conflicts due to the wording.
Re: backport proposals
Posted by Jacob Champion <ch...@gmail.com>.
On 04/27/2017 07:15 AM, Stefan Eissing wrote:
> So, the proposal could have been written as:
>
> *) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging svn merge -c
> 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387
> ^/httpd/httpd/trunk . +1: jim, ylavic
>
> Wouldn't that be easier?
+1, and I plan to write similar proposals like this when I start
backporting feature branches. They'll end up looking like
svn merge -r <branch point>:HEAD ^/httpd/httpd/branches/<branch>
--Jacob
Re: backport proposals
Posted by Yann Ylavic <yl...@gmail.com>.
On Thu, Apr 27, 2017 at 4:15 PM, Stefan Eissing
<st...@greenbytes.de> wrote:
> Take this as an observation about proposals in general, nothing wrong with this one in particular:
>
> *) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging
> trunk patch: http://svn.apache.org/r1784203
> http://svn.apache.org/r1784205
> http://svn.apache.org/r1784227
> http://svn.apache.org/r1784228
> http://svn.apache.org/r1784275
> http://svn.apache.org/r1785871
> http://svn.apache.org/r1786009
> http://svn.apache.org/r1789387
> 2.4.x patch: trunk works *after r1779573 above* (modulo CHANGES)
> ie: http://home.apache.org/~ylavic/patches/httpd2.4-hcheck-after-r1779573.patch
> FULL hcheck patch: http://home.apache.org/~jim/patches/httpd2.4-hcheck.patch
> http://svn.apache.org/r1789387
> (includes all hcheck related patches, including showstopper)
> +1: jim, ylavic
>
> So, how to check this? I tried the FULL hcheck patch. It does not work, ~50% of hunks fail.
>
> Ok, revert. Then I just tried this:
> svn merge -c 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387 ^/httpd/httpd/trunk .
>
> and all is well! So, the proposal could have been written as:
>
> *) mod_proxy_hcheck: Honor checks in Vhosts w/o hanging
> svn merge -c 1784203,1784205,1784227,1784228,1784275,1785871,1786009,1789387 ^/httpd/httpd/trunk .
> +1: jim, ylavic
>
> Wouldn't that be easier? I mean, sometimes trunk and backport may differ a lot. But most commonly, only CHANGES and message-tags need to be ignored. I myself would prefer just to copy&exec a one liner.
Agreed, I tried several merges too, that's why I added the patch I
tested in the proposal (httpd2.4-hcheck-after-r1779573.patch), which
corresponds to yours (as indicated in the trailer: "Merged
/httpd/httpd/trunk:r1784203,1784205,1784227-1784228,1784275,1785871,1786009,1789387").
Looks like it didn't help either :/