| |
| |
| |
|
Comments:
<0> Hi! I ve got a problem with the connection tracking system of the kernel, I am using 2.6.16 kernel!
<0> The prob is that if I sent a syn packet the conntrack ***umes that connection as ESTABLISHED rather than SYN_SENT and as you can image the time out is not 2minutes but 5 days. I can give an example from /proc/net/conntrack
<0> tcp 6 431977 ESTABLISHED src=192.168.1.1 dst=213.3.12.7 sport=52036 dport=80 packets=1 bytes=40 [UNREPLIED] src=213.3.12.7 dst=192.168.1.1 sport=80 dport=52036 packets=0 bytes=0 mark=0 use=1
<0> At this connection there is only one packet sent (the SYN one) with NO REPLY!, but as you can see it is ***umed as ESTABLISHED and the timeout time is almost 5 days
<0> I have asked at #iptables #netfilter, posted at forums of gentoo (that's the platform I use) but noone gave an answer, plz any help!
<1> does anyone know if the via rhine driver is still broken in the 2.6.9 branch?
<1> it appears so. the intergface never tops 100Kb/s
<2> Noted an issue with hosts glibc built with 2.4.x kernel when a 2.6.x kernel is being used. uinput (rightly so) basically ceases to be a functional module (finally figured out why, too)
<2> If anyone cares, it's the size change to ABS_MAX that tanks it
<2> *should* only be affecting Slackware users, since I think they're basically the only people left who would have 2.4.x kernel headers while using a 2.6.x kernel.
<2> Still trying to figure out whether it's worth asking Pavik or someone about
<3> Dagmar: slackware is not the only dist that has 2.4 headers
<2> It's the only one of the majors left that's in that transitional state
<3> I still run with 2.4.x headers & have on occ***ions used 2.6 kernels & I do not use slackware at all
<2> Ah
<2> You won't be able to use uinput very well with that. Drove me nuts this week
<3> givenI dont even know what winput is, it probably wont concern me a great deal
<2> The thing that inits a uinput device checks the size of the struct it's handed. With the 2.4.x headers that would amount to 604 bytes. With the 2.6.x headers, 1116.
<2> All this work for a bloody keyboard.
<2> I got a logitech g15 keyboard a few weeks ago and have been dorking around with it under Linux.
<3> never had akeyboard issue yet with 2.6 on 2.4 headers
<2> You're not trying to read it through uinput either tho
<2> The g15 has a *lot* of extra keys that it would be useful to hook that way
<2> I'm glad I figured out how to get around it before I started struggling with Device::USB. Heh
<3> well given I have zer5o usb hardware, I doubt I will be really concerned :)
<2> Well, the keyboard is nice to type on, even if it's the size of a freaking battleship
<3> usb has been & will probably be a sl.ow ****y standard thats a complete waste of time for serious computer users
<2> Imagine a full-sized 104-key with a second numpad on the left
<3> it's fine for the bells n whistles user who does stuff all with their hardware
<2> Hardly. It's really, really cheap, and not that complex to talk across
<3> really really cheap hardly makes it decent
<3> I have the same feeling towards any nic based on the 8139 cxhipset
<3> it's crap & I wouldn't touch it
<2> Using it for ethernet is not good, I agree
<3> dont know why people having difficulty with crappy standards/hardware think it's the fault of kernel programming
<2> That's a pretty bad use for it, but for bulk transfers and relatively infrequently chirpy things (like keyboards) it seems reasonable enough
<3> I'llcontinue using my obsolete hardware which works for 2.4 & 2.6 regardless of headers then & leave you too your much better cheaper hardware
<2> I haven't yet found a usb-connectable LCD display for under $100
<2> I paid $60 for the keyboard.
<3> I paid zero for my keyboard woopy doo
<3> I also paidzero for my typist & suffer for it everyday:)
<4> hi all!
<4> *g*
<5> hi
<6> hi
<6> how can solve this ?
<5> hi
<6> ide0: I/O resource 0x1F0-0x1F7 not free.
<6> ide0: ports already in use, skipping probe
<6> using debian sid kernel 2.6.16
<7> why is there no CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND or CONFIG_CPU_FREQ_DEFAULT_GOV_POWERSAVE?
<8> bah errors
<8> http://pastebin.com/728065
<8> i was trying to copile latest stable 2.6 kernek with beyond sources 3
<8> init 0
<3> What chaned cef?
<3> What changed cef?
<9> mm2 got released
<10> hm
<10> shouldn't it be like this?
<10> 2.{6.{17{-rc4,-rc4-mm2},16.16,15.6,14.5,13.4,12.6,11.12},4.{33-pre3,32},2.{27-rc2,26},0.40}
<10> ;P
<9> or what about 2.{6.{17{-rc{4,4-mm2},16.16,15.6,14.5,13.4,12.6,11.12},4.{33-pre3,32},2.{27-rc2,26},0.40}
<10> doesn't work
<9> yes it does. that expands to -rc4 and -rc4-mm2
<3> then you could just chuck the url into the topic & update the page automagically everytime something comes from lkml
<9> why should I be doing a #kernel webpage? hardly got the time to update the topic half the time
<3> :P
<9> pfffft, then why not just send them to kernel.org?
<10> doesn't work for me then :), it's correct though bash stops liking it or something :)
<3> because they're too lame too go there
<3> but all your lemming mates come in here cef :P
<10> ;p
<9> kastberg_: :p to you too
<9> gb: whatever.. I'm not in the mood. later
<3> uhuh
<11> why does it say the latest stable kernel is a patch
<12> because it was easier for them to post a patch?
<11> yea I realised you have to press f
<11> to get the source
<11> really odd stuff when configuring through the oldconfig
<11> it skips some things
Return to
#kernel
or
Go to some related
logs:
truemobile 1350 apt-get
#css
#css
#linux
Perl setuid emulation
php-cgi Accelerator gentoo
ubunto software raid
iptables frostwire
ubuntu remove oss from kernel
udev run-program exec of program failed
|
|