Added (libevent,tor) and modified glibc-compat

Added (libevent,tor) and modified glibc-compat

Visit http://libevent.org for details about libevent.
Visit https://www.torproject.org for details about Tor

Currently tor and libevent compiles fine, however
while executing them (for instance in devenv) some errors
still needs to be resolved.
Tor : geoip files and config files also need to supplied
Libevent : Bugs are still left ( for example sockets api call)

Fixes for bugs such as above mentioned will be sent over the
duration of this month, the aim of this pull request is to get
the build reviewed.

R=binji@google.com

[deepankar-tyagi, 2015-07-09 18:04:22]

On 2015/07/09 17:59:43, deepankar-tyagi wrote:

Just noticed that gedit also auto-indented the glibc-compat makefile(sources:= 
),
Let me know if it needs to be changed back.

[Sam Clegg, 2015-07-16 00:26:29]

sbc@chromium.org changed reviewers:
+ sbc@chromium.org

[Sam Clegg, 2015-07-16 00:26:30]

nice work!

just a few nits.

https://codereview.chromium.org/1226403004/diff/1/ports/glibc-compat/Makefile
File ports/glibc-compat/Makefile (right):

https://codereview.chromium.org/1226403004/diff/1/ports/glibc-compat/Makefile...
ports/glibc-compat/Makefile:44: src/writev.c
Yes, please change this indentation back

https://codereview.chromium.org/1226403004/diff/1/ports/libevent/build.sh
File ports/libevent/build.sh (right):

https://codereview.chromium.org/1226403004/diff/1/ports/libevent/build.sh#new...
ports/libevent/build.sh:1: # Copyright (c) 2015 The Native Client Authors. All
rights reserved.
nit: drop the (c)

https://codereview.chromium.org/1226403004/diff/1/ports/libevent/build.sh#new...
ports/libevent/build.sh:16: EXTRA_CONFIGURE_ARGS+=" --disable-libevent-regress "
nit: drop the trailing space

https://codereview.chromium.org/1226403004/diff/1/ports/libevent/build.sh#new...
ports/libevent/build.sh:22: # EXTRA_CONFIGURE_ARGS+=" --enable-shared=no
--enable-static=yes"
nit: indent these commented out lines too.

Do you have any reason to believe this might be required?  In not then I would
just drop the comment.

https://codereview.chromium.org/1226403004/diff/1/ports/tor/build.sh
File ports/tor/build.sh (right):

https://codereview.chromium.org/1226403004/diff/1/ports/tor/build.sh#newcode1
ports/tor/build.sh:1: # Copyright (c) 2015 The Native Client Authors. All rights
reserved.
nit: drop the (c)

https://codereview.chromium.org/1226403004/diff/1/ports/tor/build.sh#newcode27
ports/tor/build.sh:27: EXTRA_CONFIGURE_ARGS+=" --disable-gcc-hardening"
I would just set all this stuff at the global scope rather than overriding
ConfigureStep.

https://codereview.chromium.org/1226403004/diff/1/ports/tor/build.sh#newcode55
ports/tor/build.sh:55: DefaultBuildStep
Probably cleaner to do it removing the NDEBUG (see ports/grep/build.sh)

[deepankar-tyagi, 2015-07-17 17:19:08]

On 2015/07/16 00:26:30, Sam Clegg wrote:
> nice work!
Thanks.

>
https://codereview.chromium.org/1226403004/diff/1/ports/libevent/build.sh#new...
> ports/libevent/build.sh:22: # EXTRA_CONFIGURE_ARGS+=" --enable-shared=no
> --enable-static=yes"
> Do you have any reason to believe this might be required?  In not then I would
> just drop the comment.

when building with glibc, I observed that nexe were compiled with dynamic
linking.
I checked the config log and found that configure correctly detected static
linking
and passed static flag, inspite of that nexes were not statically linked.
I deprioritized this issue for a while, right now my priority is to get libevent
and 
tor working at least with newlib and clang-newlib
> 
> https://codereview.chromium.org/1226403004/diff/1/ports/tor/build.sh#newcode27
> ports/tor/build.sh:27: EXTRA_CONFIGURE_ARGS+=" --disable-gcc-hardening"
> I would just set all this stuff at the global scope rather than overriding
> ConfigureStep.
Did you meant something like this 
EXTRA_CONFIGURE_ARGS=" --disable-tool-name-check --disable-system-torrc
--disable-gcc-hardening"
on top of build.sh

[Sam Clegg, 2015-07-17 19:32:40]

On 2015/07/17 17:19:08, deepankar-tyagi wrote:
> On 2015/07/16 00:26:30, Sam Clegg wrote:
> > nice work!
> Thanks.
> 
> >
>
https://codereview.chromium.org/1226403004/diff/1/ports/libevent/build.sh#new...
> > ports/libevent/build.sh:22: # EXTRA_CONFIGURE_ARGS+=" --enable-shared=no
> > --enable-static=yes"
> > Do you have any reason to believe this might be required?  In not then I
would
> > just drop the comment.
> 
> when building with glibc, I observed that nexe were compiled with dynamic
> linking.
> I checked the config log and found that configure correctly detected static
> linking
> and passed static flag, inspite of that nexes were not statically linked.
> I deprioritized this issue for a while, right now my priority is to get
libevent
> and 
> tor working at least with newlib and clang-newlib
> > 
> >
https://codereview.chromium.org/1226403004/diff/1/ports/tor/build.sh#newcode27
> > ports/tor/build.sh:27: EXTRA_CONFIGURE_ARGS+=" --disable-gcc-hardening"
> > I would just set all this stuff at the global scope rather than overriding
> > ConfigureStep.
> Did you meant something like this 
> EXTRA_CONFIGURE_ARGS=" --disable-tool-name-check --disable-system-torrc
> --disable-gcc-hardening"
> on top of build.sh

Yes.

[Sam Clegg, 2015-08-18 05:26:22]

lgtm