[SOLVED] Error in prepare() when building kernel linux-omap

[abhay]

Hello

I wanted to rebuild the kernel for Pandaboard so cloned the github repository on the Pandaboard, changed directory to linux-omap. I uncommented the lines for menuconfig and the return 1. Thats it, and ran makepkg -Acs --asroot.

Everything was downloaded and then after several patches being applied, I received the following error:

$this->bbcode_second_pass_code('', '
:
:
patching file security/commoncap.c
patching file security/device_cgroup.c
patching file security/security.c
==> ERROR: A failure occurred in prepare().
Aborting...
')

What could this error be? Is some patch failing? How can I find which one? Can't find a build log file. Is there one?

Thank you for your help.

[abhay]

I redid the whole process and got some more info:

$this->bbcode_second_pass_code('', '
:
:
patching file drivers/net/wireless/ti/wl12xx/wl12xx.h
Hunk #1 FAILED at 30.
1 out of 1 hunk FAILED -- saving rejects to file drivers/net/wireless/ti/wl12xx/wl12xx.h.rej
patching file drivers/net/wireless/ti/wl18xx/main.c
patching file drivers/net/wireless/ti/wlcore/boot.c
Hunk #1 FAILED at 84.
1 out of 1 hunk FAILED -- saving rejects to file drivers/net/wireless/ti/wlcore/boot.c.rej
patching file drivers/net/wireless/ti/wlcore/wlcore.h
Hunk #2 succeeded at 435 (offset -3 lines).
==> ERROR: A failure occurred in prepare().
Aborting...
')

[abhay]

Ok. so i manually checked those patch and rej files and applied the said patches. there were a couple of files that were quite different than expected in the patch.

Why should that be? Isn't the patch built against the 3.7 kernel source? Are all the patches still valid and necessary for the linux-omap kernel?

Unfortunately the linux-armv7h branch doesn't have the board-omap4panda.c file, so I'm assuming that that branch is not to be used for the Pandaboard? Does it mean that the Pandaboard will not get kernel updates henceforth?

@WarHeadSE and others, could you please confirm? Thanks in advance.

[kmihelich]

You need to check what you're doing because there is absolutely no problem with with PKGBUILD or the patches.

[abhay]

Here are the EXACT steps I follow: (I followed them once again to verify before sending you this reply)

$this->bbcode_second_pass_code('', '
git clone git://github.com/archlinuxarm/PKGBUILDs.git
cd PKGBUILDs
git checkout core/linux-omap
cd core/linux-omap
makepkg -Acs
')

Result:
$this->bbcode_second_pass_code('', '
:
:
patching file drivers/net/wireless/ti/wl12xx/wl12xx.h
Hunk #1 FAILED at 30.
1 out of 1 hunk FAILED -- saving rejects to file drivers/net/wireless/ti/wl12xx/wl12xx.h.rej
patching file drivers/net/wireless/ti/wl18xx/main.c
patching file drivers/net/wireless/ti/wlcore/boot.c
Hunk #1 FAILED at 84.
1 out of 1 hunk FAILED -- saving rejects to file drivers/net/wireless/ti/wlcore/boot.c.rej
patching file drivers/net/wireless/ti/wlcore/wlcore.h
Hunk #2 succeeded at 435 (offset -3 lines).
==> ERROR: A failure occurred in prepare().
Aborting...
')

If you have 10 minutes, could you please reproduce the above steps on a Pandaboard if you have one? I will really appreciate it.

Thanks,

Abhay

[kmihelich]

I'm not in the habit of posting broken patches or broken PKGBUILDs, and your insinuation that I would do so, or not bother to verify yet again that the PKGBUILD and sources I committed to git (which built the kernel in the repo) are still working is infuriating.

Once again, there is nothing wrong with the PKGBUILD or the patches, you need to check what you're doing. Specifically, make sure you're not applying the patches to an already existing source extraction that is already patched.

[abhay]

Kevin,

The same steps in core/linux-armv7 work absolutely fine.

Abhay

[abhay]

It may not be a habit, but there certainly can be a mistake made from time to time. Why are you taking this personally? Nobody is insinuating anything. Maybe it wasn't you, maybe it was someone else who made a mistake? Does it hurt to check?

[kmihelich]

Please actually read what I'm writing.

I did check, twice now, in two different scenarios, and the PKGBUILD works just fine and the patches apply. The problem is on your end.

[abhay]

OK. In that case, can you give me some pointers how I can debug the problem at my end?

As I mentioned, I started with a fresh ALARM image and in a brand new directory and followed the exact 5 steps I mentioned before.

What could I be doing wrong?