Version: 3.3 (using KDE 3.3.0, (3.1)) Compiler: gcc version 3.3.4 (Debian 1:3.3.4-12) OS: Linux (i686) release 2.6.10-rc1-mm2 top output shows: PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 3889 root 15 0 401m 185m 1112 S 55.6 36.9 117:18.46 XFree86 4905 Amit 15 0 242m 91m 1904 S 0.0 18.1 5:29.89 nspluginviewer 4619 Amit 15 0 205m 76m 2004 S 0.3 15.2 5:02.13 nspluginviewer I have 3 konqueror windows open with several tabs in them. My system is up for just 2 days now but there's just 41M free of 512M and swap is 0M free of 512M. I just can't keep this system on for > 5 days. This time it's been particularly bad. This is a very nasty bug for me; I couldn't even run a process because there was no available memory for 'fork'.
On Saturday 06 November 2004 02:30, Amit Shah wrote: > top output shows: > > PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND > 3889 root 15 0 401m 185m 1112 S 55.6 36.9 117:18.46 XFree86 > 4905 Amit 15 0 242m 91m 1904 S 0.0 18.1 5:29.89 nspluginviewer > 4619 Amit 15 0 205m 76m 2004 S 0.3 15.2 5:02.13 nspluginviewer > > > I have 3 konqueror windows open with several tabs in them. My system is up > for just 2 days now but there's just 41M free of 512M and swap is 0M free > of 512M. I just can't keep this system on for > 5 days. This time it's been > particularly bad. > > This is a very nasty bug for me; I couldn't even run a process because > there was no available memory for 'fork'. It's rather certain that this is a leak in flash player or some other plugin, not in nspluginviewer. Those plugins are known to be very buggy. Do you have a testcase that we can use to confirm this?
> ------- Additional Comments From staikos kde org 2004-11-06 12:36 ------- > Do > you have a testcase that we can use to confirm this? > I have a few tabs with some arstechnica windows open. They have flash on their sites. Some of the links are: http://arstechnica.com/articles/paedia/hardware/pcie.ars/3 http://arstechnica.com/articles/paedia/cpu/pentium-1.ars/1 http://arstechnica.com/articles/paedia/cpu/pentium-2.ars/1 http://arstechnica.com/articles/paedia/cpu/ppc-1.ars/1 I've currently disabled the shockwave plugin; I'll check the RAM usage after a couple of days. Hopefully, the situation would be better.
Yeah; back after the weekend and It's using 60M of the swap. This is much better (earlier, it would've been something like 75% of swap usage -- total is 512M).
We can't fix flash's bugs.