Bug 175500 – sys-apps/915resolution ~x86-fbsd ebuild/patch & keyword request

• Status: VERIFIED FIXED
• Product: Gentoo/Alt

[Description Joe Peterson (RETIRED) 2007-04-21 21:27:40 UTC]

I've ported sys-apps/915resolution to Gentoo/FreeBSD, and attached are the new ebuild files.  I used similar methods and some code from a BSD port of 855resolution found here: http://www.sigusr1.org/~kurahone/misc-hacks/.

Now people with the Intel i810/i855/i915 (etc.) chips should be able to patch their BIOS and get wide-screen resolution like 1280x768.

[Comment 1 Joe Peterson (RETIRED) 2007-04-21 21:28:08 UTC]

Created attachment 116943 [details]
ChangeLog

[Comment 2 Joe Peterson (RETIRED) 2007-04-21 21:28:23 UTC]

Created attachment 116944 [details]
915resolution-0.5.3-r1.ebuild

[Comment 3 Joe Peterson (RETIRED) 2007-04-21 21:28:43 UTC]

Created attachment 116945 [details]
files/915resolution-0.5.3-freebsd.patch

[Comment 4 Joe Peterson (RETIRED) 2007-04-21 21:29:07 UTC]

Created attachment 116946 [details]
files/initd-r1

[Comment 5 Joe Peterson (RETIRED) 2007-04-21 22:09:39 UTC]

One note: -O2 breaks this build (the assembler part), and it's not really needed anyway.  I set CFLAGS to "" in the ebuild, which I have just heard is not recommended (makes sense!).  So really, we should filter the -O2 flag out instead...

[Comment 6 Joe Peterson (RETIRED) 2007-04-22 00:34:36 UTC]

Created attachment 116960 [details]
915resolution-0.5.3.ebuild

Now use flag-o-matic to remove various flags, since optimization breaks the assembly part of this build.  This is more correct than hard-setting CFLAGS.

[Comment 7 Joe Peterson (RETIRED) 2007-05-02 17:49:30 UTC]

Created attachment 117972 [details]
files/initd-r1

Fixed issue where returned error codes (from 915resolution program) will not be remembered if not on last mode tried.  Solution is to remember the last *bad* (non-zero) error code returned and return that code.  Please advise if it would be better to return a fixed error code if one non-zero code is returned.

Also now insert separators between multiple modes in the log file for readability.

[Comment 8 Joe Peterson (RETIRED) 2007-05-13 19:18:20 UTC]

Comment on attachment 116960 [details]
915resolution-0.5.3.ebuild

Changing rev back to original (no -r1)

[Comment 9 Joe Peterson (RETIRED) 2007-05-13 19:18:54 UTC]

Summary change

[Comment 10 Joe Peterson (RETIRED) 2007-05-13 19:20:45 UTC]

Note: new initd file is named initd-r1 in new ebuild and in attachment.  I have changed the name of the ebuild back to reflect the old version number, so I am unsure how to name this initd file; leaving as-is for now; please change as appropriate when committing...

[Comment 11 Joe Peterson (RETIRED) 2007-06-08 17:59:59 UTC]

Created attachment 121509 [details]
915resolution-0.5.3-r1.ebuild

Mods to ebuild (and version bump)

[Comment 12 Joe Peterson (RETIRED) 2007-06-08 18:02:56 UTC]

Comment on attachment 116943 [details]
ChangeLog

Ignore previously attached ChangeLog

[Comment 13 Joe Peterson (RETIRED) 2007-06-08 18:06:20 UTC]

Port to ~x86-fbsd, bump version, keyword this arch only, fix init.d script.

Note to maintainer: please test on other archs and then re-add their keywords (thanks!)

[Comment 14 Joe Peterson (RETIRED) 2007-06-09 19:21:55 UTC]

Re-opening, so we don't lose track that maintainer should test on other archs...

[Comment 15 Patrick McLean 2007-06-11 02:13:29 UTC]

Re-keyworded for x86, adding amd64 to CC as I don't have an amd64 machine with an intel graphics card to test with

[Comment 16 Christoph Mende (RETIRED) 2007-06-21 19:43:52 UTC]

amd64 done

[Comment 17 Patrick McLean 2007-07-10 18:32:59 UTC]

Should be able to close this now...

[Comment 18 Joe Peterson (RETIRED) 2007-07-10 20:59:53 UTC]

OK, closing.  Thanks, everyone!