login
Search
Linus Torvalds announced 2.6.0-test11, calling it the "Beaver in Detox" kernel and saying it was "for everybody who thought 'stoned beaver' wasn't an appropriate name for a kernel". As to the actual changes, Linus explains:
"This is mainly brought on by the fact that the old aic7xxx driver was broken in -test10, and Ingo [interview] found this really evil test program that showed an error case in do_fork() that we had never handled right. Well, duh! While at it, this also pulls in some firewire fixes and a few potential skbuff leakage points."
Linus went on to note that he'll be too busy eating for the next few days to check his email, and from here on out the 2.6 kernel is in Andrew Morton's [interview] able hands. Whether Andrew will go on to release a 2.6.0-test12, or the official 2.6.0, this is yet to be seen.
From: Linus Torvalds [email blocked]
To: Kernel Mailing List [email blocked]
Subject: Beaver in Detox!
Date: Wed, 26 Nov 2003 12:55:00 -0800 (PST)
Ok,
for everybody who thought "stoned beaver" wasn't an appropriate name for
a kernel (yeah, I'm sure IBM really minds my naming scheme, and is deathly
afraid it will scare away customers), I'm happy to tell you that the
beaver just went into detox, and I'm taking the Thanksgiving weekend off.
I give you "Beaver in Detox", aka linux-2.6.0-test11. This is mainly
brought on by the fact that the old aic7xxx driver was broken in -test10,
and Ingo found this really evil test program that showed an error case in
do_fork() that we had never handled right. Well, duh!
While at it, this also pulls in some firewire fixes and a few potential
skbuff leakage points.
Please don't even bother sending me patches, because I'll be stuffing my
face away from email over the next few days. And after that it will be up
to Andrew to say how to go on from here.
Mmmm. Turkey.
Linus
Summary of changes from v2.6.0-test10 to v2.6.0-test11
============================================
Bart De Schuymer:
o [BRIDGE]: Fix netfilter config tests
Ben Collins:
o Lastminute IEEE-1394 fixes
Benjamin Herrenschmidt:
o [libata] Fix flush of Device Control register to device
Dave Kleikamp:
o JFS: Avoid segfault when dirty inodes are written on readonly mount
David Mosberger:
o ia64: Correct FIXADDR_USER_END so that single-stepping in the gate
DSO works again
o ia64: Make core-dumps work even when executing in the gate DSO
David S. Miller:
o [NET]: In sock_queue_rcv_skb(), do not deref skb->len after it is
queued to the socket
o [PPPOE]: Do not leak SKB if sock_queue_rcv_skb() fails
o [ECONET]: Do not leak SKBs if ec_queue_packet() fails
David Stevens:
o [IPV6]: Multicast output bypasses netfilter hooks, fix
Hideaki Yoshifuji:
o [IPV6]: Redo stateless addrconf properly
James Bottomley:
o Updated state model for SCSI devices
o Fix locking problems in scsi_report_bus_reset() causing aic7xxx to
hang
Jan Marek:
o fix smsc-ircc2.c double free
Jean Tourrilhes:
o [IRDA]: Fix SKB leaks
Jeff Garzik:
o [libata] bump versions for core and serverworks driver
Linus Torvalds:
o Fix error return on concurrent fork() with threaded exit()
Mike Anderson:
o scsi device ref count (update)
Pekka Pietikäinen:
o [libata] add Promise PCI id
Pete Zaitcev:
o [SPARC]: Add stub sched_clock() implementation
o [SPARC]: Fix build failures in IGA frame buffer introduced by
janitor changes
Stephen Hemminger:
o prevent oops from read of proc entry for tty drivers
From: John Cherry [email blocked]
Subject: Re: Beaver in Detox! (compile stats)
Date: Wed, 26 Nov 2003 16:20:11 -0800
Linux 2.6 Compile Statistics (gcc 3.2.2)
Warnings/Errors Summary
Kernel bzImage bzImage bzImage modules bzImage modules
(defconfig) (allno) (allyes) (allyes) (allmod) (allmod)
----------- ----------- -------- -------- -------- -------- ---------
2.6.0-test11 0w/0e 0w/0e 170w/ 0e 12w/0e 3w/0e 209w/0e
2.6.0-test10 0w/0e 0w/0e 170w/ 0e 12w/0e 3w/0e 209w/0e
2.6.0-test9 0w/0e 0w/0e 174w/ 0e 12w/0e 3w/0e 217w/0e
2.6.0-test8 0w/0e 0w/0e 178w/ 0e 12w/0e 3w/0e 219w/0e
2.6.0-test7 0w/0e 0w/0e 173w/ 1e 8w/0e 3w/0e 226w/0e
2.6.0-test6 0w/0e 1w/0e 188w/ 1e 12w/0e 3w/0e 260w/2e
2.6.0-test5 0w/0e 2w/0e 205w/ 9e 15w/1e 0w/0e 305w/5e
2.6.0-test4 0w/0e 2w/0e 797w/55e 68w/1e 3w/0e 1016w/34e
2.6.0-test3 0w/0e 2w/0e 755w/66e 62w/1e 7w/9e 984w/42e
2.6.0-test2 0w/0e 1w/0e 952w/65e 63w/2e 7w/9e 1201w/43e
2.6.0-test1 0w/0e 1w/0e 1016w/60e 75w/1e 8w/9e 1319w/38e
Web page with links to complete details:
http://developer.osdl.org/cherry/compile/
Daily compiles (ia32):
http://developer.osdl.org/cherry/compile/2.6/linus-tree/running.txt
Daily compiles (ia64):
http://developer.osdl.org/cherry/compile/2.6/linus-tree/running64.txt
Latest changes in Linus' bitkeeper tree:
http://linux.bkbits.net:8080/linux-2.5
John
P.S. All compile data should be posted by 5PM PST. It is still
building....
Hee hee hee
Nice to see that the same humor which appeared when 2.4.0 was released has resurfaced here :)
Turkey rocks, BTW. I wonder if OSDL feel like turkeys, though.....
Well..
Let just hope we don't get a greased turkey release for 2.6. :)
Naming
Well, erm, surely it should have been the 'Beaver - Cold Turkey' release then. Duh!
I like poo
Yay! Poo r0x!
Re: I like poo
And so the Slashdotters begin...
The Problem w/ Being No. 1
The problem with being the #1 kernel news site is. . .
work to do, before a 100% stable/functional kernel
I just placed new updates, from 2.6.0-test10-mm1, of:
Documentation/should-fix.txt
Documentation/must-fix.txt
wow, quite a list, what are t
wow, quite a list, what are the chances of all this getting fixed before the stable release? (not that i could complain since 2.5.45 or so this kernel runs quite fine for me, yeah even with preemt enabled :)
Performance problems with AS?
Are the performance problems with the anticipatory scheduler fixed? I remember there was some stuff about very poor performance starting with test4 or something like that. No mention of that in those lists given in the parent post at least, so I guess they are fixed?
Re:
Yes
aha152x broken in 2.6.0-test11
It looks like this particular scsi driver is destined for the scrap heap. A search shows problems with it since 2.5 began.
speedtouch unselectable?
Dunno if you can reproduce this: there should be a speedtouch usb driver (in drivers/usb/misc) which does not appear in menuconfig or xconfig! I can not select it... quite bad, as it is my modem and i can only use it under 2.4 (user mode, which doesn't work in 2.6).
So, there is a speedtch.c in the sources but no option in the menuconfig. anyone else?
Speedtouch
You must have USB and ATM enabled for it to appear in the menu
speedtouch and Kconfig
If you know where a certain module/function is in the source tree you can check the Kconfig file in that directory.
In this case speedtch.c is in drivers/usb/misc. Looking at the
Kconfig in that directory you see that:
config USB_SPEEDTOUCH tristate "Alcatel Speedtouch USB support" depends on USB && ATMPeder
You need to turn on ATM and U
You need to turn on ATM and USB support.
ATM support is still experimental, so you'll need
"Prompt for development and/or incomplete code/drivers"
as well.
via apic irq routing
Whats the deal with via apic irq routing? Is this a won't fix bug?
Re: via apic irq routing
well, I've ever had via APIC irq routing problems with both 2.4 and 2.5/6, but they've been gone with test6 or 7 - not 100% sure about what version it exactly was, was using patches as of test3... - (and with 2.4.22), so the list's maybe a little bit outdated. Of course, there are still problems left, especially when I try to use IO APIC without ACPI.
btw. this is on a EPOX 8k5a3+ (KT 333).
I have the same problem
I noticed it when the via-rhine network driver wouldn't work with ACPI enabled. (2.6.0-test9 and 2.6.0-test10) Turning ACPI off fixed the problem, but I found the solution to be pretty non-intuitive.
Re: I have the same problem
did you try acpi + IO apic ? this combination works well here. (epox 8k5a3+, via-rhine onboard)
preemption
did they fix preemption?
I can't see it in the must or should fix list.
QnX
What do you mean fix? It's been working for quite a while... You mean stability?
I read of problems with preem
I read of problems with preemption enabled.
So I think that's a bug to be fixed.
It's nit mentioned in the tofix lists.
Even Linus said: don't use test10 with preempt enable.
No real preempt problem
As far as I know the problem didn't exist - one of the last mails from the thread:
still problems
There still are problems. With test9, test10 and test10-mm1 I cannot start X when preeemt is compiled in. Without preeemt I have no problems at all.
The difficulty seems to be that the problems are not always reproducible and seem not related to a specific combination of drivers/hardware/settings (except for preemt then).
bye,
pieter
yes, the pre-emption problem still exists
I'm running -test11 now, and it's great, but I have pre-emption disabled.
If I enable pre-emption, after a random number of hours (usually 2 - 10), my cpu and disk go into overdrive, ram starts rising and the load goes so high that I can't even open a tty. I have no option but to hold down the power button (while the disk is audibly active).
Ciaran O'Riordan
/me = QnX Now it sounds reso
/me = QnX
Now it sounds resonabely! I read about ppl having problems while booting and such, but I've had preempt enabled since.... a long time back, and it's always been working for me. However, yesterday, after about 5h uptime I noticed that X gained 99.9% CPU usage (didn't noticed it until dcgui-qt started to hang, by that time X was taking large load for over 20 minutes, so god damn that O(1) scheduler works well in 2.6 ;)). Couldn't do anything else than pkill X . But I never thought it might have been preempt's fault.. But now it looks possible... Don't you think?
RE: yes, the pre-emption problem still exists
Yeah I have the same exact problem but if I let it go after a while the disk usage stops and ram use goes back down to normal.
Oughtn't their be NO warnings at all when compiling?
I'd have thought that generally when there's something like 150 or so warnings you're software is probably kinda fragile.
What's worse...
AFAIKT "make allyesconfg" only compiles the files that are known to compile cleanly.
Adaptecs dpt_i2o still does not compile at all for instance
(as noted in the must_fix along with lots of other drivers).
The warnings are often pretty harmless
Peder
That doesn't make sense
That doesn't make sense. If it's harmless then there's no need to issue a warning.
Compilers do this
The GNU C compiler (like most) often issues warnings along the lines
of "foo might be problem" when it can't completely analyze a particular situation. A human can look at the code and determine that the problem
doesn't actually exist. In theory, one can re-arrange the code to
make it easier to analyze and avoid the bogus warning; in practice
doing so might have negative effects far worse than the warning.
Also
Also, one must see that the linux kernel is designed to run on quite a lot of platforms, both 32 and 64 bit. Thus, a programmer may write code that makes a difference on one platform, but doesn't on the other, so on the other, the compiler will likely warn. 'comparison is always false due to limited range of data type' is one of those warnings.
Works well
My beaver is in detox now, and it works pretty well. No problems yet. The only thing that bothers me is that the beaver only includes ALSA 0.9.7, so I'm down from 1.0.0-pre3 with the 2.4 kernels. Anyway, I assume that, now that ALSA is a part of the official kernel, the 2.6 kernels will keep pace with it soon.
can the number of warnings get to 0 ?
AS arch/i386/boot/setup.o
arch/i386/boot/setup.S: Assembler messages:
arch/i386/boot/setup.S:165: Warning: value 0x37ffffff truncated to 0x37ffffff
a bit of a stupid warning since the truncated value is the same as the original
chris (newbie)
P.S pre-emptive was turned on
AIC7xxx still broken
"...brought on by the fact that the old aic7xxx driver was broken in -test10"
It's still broken in test11 for me. I'm getting the kernel panic unable to mount root FS problem. Back to 2.4...
meta@pobox.com
ide-scsci
Hi all,
Will 2.6 final eventually have ide-scsi? I'm using -test11 from http://people.redhat.com/arjanv/2.5/ and i'm wondering if arjanv will pack the next rpm with ide-scsi. I've been using /dev/hdc as mentioned in the README but still i've problem getting my cdwriter/external floppy/digi cam to work. 2.6 really does improve my laptop performance and i'm thinking a way to fix all of this so i can stick to my laptop.