Bug List: (This bug is not in your last search results)   Show last search results      Search page      Enter new bug
Bug#: 171529
Alias:
Product:
Component:
Status: RESOLVED
Resolution: FIXED
Assigned To: Steev Klimaszewski <steev@gentoo.org>
Hardware:
OS:
Version:
Priority:
Severity:
Reporter: Gordon Malm <gengor@gentoo.org>
Add CC:
CC:
Remove selected CCs
URL:
Summary:
Status Whiteboard:
Keywords:

Filename Description Type Creator Created Size Actions
madwifi-ng-0.9.3.ebuild madwifi-ng-0.9.3.ebuild text/plain Gordon Malm 2007-03-20 07:40 0000 2.75 KB Details
madwifi-ng-0.9.3-uudecode-gcda-fix.patch madwifi-ng-0.9.3-uudecode-gcda-fix.patch patch Gordon Malm 2007-03-20 07:40 0000 333 bytes Details | Diff
madwifi-ng-0.9.3.ebuild-rev02 madwifi-ng-0.9.3.ebuild-rev02 text/plain Gordon Malm 2007-03-20 08:02 0000 2.73 KB Details
madwifi-build_all_ratemods.patch madwifi-build_all_ratemods.patch patch Gordon Malm 2007-03-20 15:15 0000 3.06 KB Details | Diff
madwifi-build_all_ratemods-rev2.patch madwifi-build_all_ratemods-rev2.patch patch Gordon Malm 2007-03-20 19:09 0000 3.02 KB Details | Diff
madwifi-ng-0.9.3-r1-to-r2.patch Fixes issues in comment #10 patch Gordon Malm 2007-03-21 15:02 0000 783 bytes Details | Diff
Create a New Attachment (proposed patch, testcase, etc.) View All

Bug 171529 depends on: Show dependency tree
Bug 171529 blocks:
Votes: 0    Show votes for this bug    Vote for this bug

Additional Comments: (this is where you put emerge --info)


Not eligible to see or edit group visibility for this bug.






View Bug Activity   |   Format For Printing   |   XML   |   Clone This Bug


Description:   Opened: 2007-03-20 07:39 0000
This is my first attempt at ebuild writing other than patching small bits so
please review/test.  It is based off the madwifi-ng-0.9.2 ebuild.

The madwifi-ng-tools-0.9.2 ebuild can simply be copied to
madwifi-ng-tools-0.9.3 I believe.  Atleast that is what I did for the moment.

Compiles, installs and works on x86/hardened hostapd machine.  Clients have
connectivity.  Very little testing otherwise.  Have not tried it with
wpa_supplicant yet, too late, need sleep.

Reproducible: Always

------- Comment #1 From Gordon Malm 2007-03-20 07:40:33 0000 -------
Created an attachment (id=113835) [details]
madwifi-ng-0.9.3.ebuild

------- Comment #2 From Gordon Malm 2007-03-20 07:40:56 0000 -------
Created an attachment (id=113836) [details]
madwifi-ng-0.9.3-uudecode-gcda-fix.patch

------- Comment #3 From Gordon Malm 2007-03-20 08:02:51 0000 -------
Created an attachment (id=113838) [details]
madwifi-ng-0.9.3.ebuild-rev02

revised ebuild (obsoletes madwifi-ng-0.9.3.ebuild attachment):
-testing (~) not stable
-don't export variables

------- Comment #4 From Steev Klimaszewski 2007-03-20 14:05:28 0000 -------
Very nice - the ARCH bit (well, really the src_compile() isn't needed at all)
isn't necessary (that I have seen) - I actually bumped this last night
sometime, and didn't get this bug until this morning.  Great work though, and
look forward to hearing from you if there are any bugs (aside from a few known
ones ;) )

------- Comment #5 From Gordon Malm 2007-03-20 15:14:18 0000 -------
One issue.  With madwifi-ng-0.9.3 rate control module selection is done at
module load time instead of compile time as in madwifi-ng-0.9.2*.  No special
USE flags or any related stuff is needed.  Patch against the
madwifi-ng-0.9.3.ebuild in portage incoming now.

Regarding the ARCH stuff: You are correct, I get rid of src_compile() and no
problems.  I was getting ARCH mismatch errors related to madwifi's ARCH
detection script, scripts/get_arch.mk while I was working on this but not
anymore. *shrug*

------- Comment #6 From Gordon Malm 2007-03-20 15:15:08 0000 -------
Created an attachment (id=113878) [details]
madwifi-build_all_ratemods.patch

------- Comment #7 From Steev Klimaszewski 2007-03-20 16:48:28 0000 -------
Good catch - when I get out of work, I will apply and bump

------- Comment #8 From Gordon Malm 2007-03-20 19:09:49 0000 -------
Created an attachment (id=113910) [details]
madwifi-build_all_ratemods-rev2.patch

BUILD_PARAMS variable is also no longer used.  I missed removing that in the
first patch.

------- Comment #9 From Steev Klimaszewski 2007-03-21 00:24:10 0000 -------
Bumped and added to CVS - Thanks for the work on this :)

------- Comment #10 From Daniel Drake 2007-03-21 13:46:50 0000 -------
Some issues with the latest ebuild:

BUILD_PARAMS has been removed, now madwifi builds against the running kernel
rather than the one in /usr/src/linux and this is not the correct behaviour

IUSE is missing from the ebuild

Using ${PN}-${PV} is redundant, use ${P} instead (in the epatch line)

------- Comment #11 From Gordon Malm 2007-03-21 15:02:33 0000 -------
Created an attachment (id=113972) [details]
Fixes issues in comment #10

Those are my fault entirely.  Sorry, still learning.  Here is a patch against
-r1.  Thank you for the constructive feedback.

------- Comment #12 From Daniel Drake 2007-03-21 15:43:21 0000 -------
Those changes look correct, thanks!

------- Comment #13 From Steev Klimaszewski 2007-03-21 15:53:14 0000 -------
Fixed, I hope.  Added IUSE after it was pointed out that it is a MUST
apparently.  Repoman didn't complain, so I assumed it was correct.  If I keep
screwing up the ebuild it will never get stabled ;)

Bug List: (This bug is not in your last search results)   Show last search results      Search page      Enter new bug