Gentoo Archives: gentoo-amd64

From: Duncan <1i5t5.duncan@×××.net>
To: gentoo-amd64@l.g.o
Subject: [gentoo-amd64] Re: Kernel 2.6.26 and the RTC
Date: Sat, 06 Sep 2008 00:25:24
Message-Id: pan.2008.09.06.00.25.10@cox.net
In Reply to: Re: [gentoo-amd64] Kernel 2.6.26 and the RTC by Peter Humphrey
1 Peter Humphrey <peter@××××××××××××××.org> posted
2 200809051612.09917.peter@××××××××××××××.org, excerpted below, on Fri, 05
3 Sep 2008 16:12:09 +0100:
4
5 > I've had RTC compiled into my kernels all along - still have. I don't
6 > use oldconfig to create new kernels, only to see what has changed. (I
7 > usually copy the .config into the new kernel source tree, and make
8 > oldconfig to see what I need to set with menuconfig; then I make
9 > mrproper, re-copy the .config, make menuconfig and only then make &&
10 > make modules_install.)
11
12 FWIW, I use make oldconfig, but start with the rcs, so by the time of the
13 general new kernel release, I've worked out all the potential issues with
14 the new options.
15
16 ... Which brings this up. I've yet to be able to successfully compile a
17 2.6.27-rc since -rc2 (which compiled but didn't run, not sure if it was
18 new options or defective rc). I keep getting errors, having tried every
19 one since then. I filed a (kernel.org) kernel bug, but it was closed as
20 works for them, the config I posted compiles fine there. They suggested
21 my kernel tree was screwed up, but I don't see how that could be because
22 (1) I've tried with multiple rcs so it's not just a one-off, and (2) it
23 can't be the bunzip2 process screwing up as I'm continuing to do routine
24 Gentoo updates with FEATURES=buildpkg, which bzip2s the package then
25 bunzip2s it for installation, and if bzip2 was defective, I'd have
26 certainly been getting errors there.
27
28 So now I'm frustrated, but haven't yet had time to look into it. But my
29 days off this week are coming up so hopefully I will. Given that the
30 config compiles for others, I'm suspecting it may have to do with
31 makeopts and parallel make, since I run -j (unlimited jobs) for the
32 kernel make, and get hundreds of jobs going. They moved a bunch of
33 header files around for .27 and I'm guessing they don't have the make
34 dependencies quite right yet. Once I do a bit more testing I'll reopen
35 the bug with the results of a serial make, which I expect to succeed, to
36 compare against the parallel make.
37
38 --
39 Duncan - List replies preferred. No HTML msgs.
40 "Every nonfree program has a lord, a master --
41 and if you use the program, he is your master." Richard Stallman