Skip to contentSkip to conversation

patch-2.6.9 against 2.6.8.1

DISCUSSION T. Weyergraf 3 participants 3 messages Oct 19, 2004 – Oct 22, 2004

lore
Share & tools
lore
#0 lore raw

Hi all,

I just build 2.6.9 using the patch-2.6.9, as always. Previously, I was using 2.6.8.1 and i expected patch-2.6.9 to work on the 2.6.8.1 tree.
The patch-2.6.9 is somewhat "confused". Against 2.6.8.1, it fails to change the SUBLEVEL field:

<snip>
@@ -1,6 +1,6 @@
VERSION = 2
PATCHLEVEL = 6
-SUBLEVEL = 8
+SUBLEVEL = 9
EXTRAVERSION =
NAME=Zonked Quokka
</snip>

As one can see, the patch failes, since in 2.6.8.1, EXTRAVERSION is set to 1. Also, patch-2.6.9 contains the small fixes in fs/nfs/file.c, that were given in patch-2.6.8.1.

Is this the desired behaviour ? Based on the past, i expected new patches to go against the latest stable kernel ( which is reported 2.6.8.1 by kernel.org ). Will - in the future - new patches skip the 4-digit kernelpatches ?

I do not intent to start a flamewar over whether 4-digit kernelreleases are the right/wrong way to go. I am just looking for a consistent behaviour, which is at this point only given, if you expect new kernelreleases to go against the last 3-digit release.

Is that so ?

Regards,
Thomas Weyergraf

Signature · 3 lines
Thomas Weyergraf                                     [email protected]
Funny IA64 Opcode Dept: ( see arch/ia64/lib/memset.S )
"br.ret.spnt.few" - got back from getting beer, did not spend a lot.
#1 lore raw
Quoted 5 lines · T. Weyergraf

Hi all,

I just build 2.6.9 using the patch-2.6.9, as always. Previously, I was using 2.6.8.1 and i expected patch-2.6.9 to work on the 2.6.8.1 tree.

The patch is against 2.6.8.

Cheers,
Buddy

#2 lore raw
In reply to T. Weyergraf
Quoted 5 lines · T. Weyergraf

Hi all,

I just build 2.6.9 using the patch-2.6.9, as always. Previously, I was using 2.6.8.1 and i expected patch-2.6.9 to work on the 2.6.8.1 tree.

No, it applies to 2.6.8. The situation could arise that a 2.6.8.2 is released _after_ 2.6.9 and then having the 2.6.9 patch based on 2.6.8.1 would be quite confusing.
Read all of this thread for the full story:
http://www.ussg.iu.edu/hypermail/linux/kernel/0408.3/0293.html

--
Jesper Juhl