Slow xterm response
Bill Moseley
moseley at hank.org
Wed Dec 5 17:36:40 PST 2007
On Wed, Dec 05, 2007 at 08:43:13PM +0000, James Cloos wrote:
> >>>>> "Bill" == Bill Moseley <moseley at hank.org> writes:
>
> Bill> ... on my desktop I ssh into the laptop and start up a development
> Bill> web application.
>
> Bill> As soon as the laptop's screen goes blank my xterm sessions show
> Bill> latency. If I hold down a key and repeat a letter it goes along in
> Bill> fits and starts.
>
> You wrote elsewhere in the thread that the blanking isn't caused by X
> dpms but [I presume] by the laptop's firmware and/or kernel.
Well, I'm not so sure. There's dpms which disables the display but
there's also the standard screen blanking that happens at 10 minutes.
I just assumed that was X clearing the screen (the LCD backlight is
still on).
That's when the latency starts. Then if I force the display off after
the screen is blanked the latency goes away.
If I change the timeout to 5 minutes then the latency starts at 5
minutes. So, it seems to be triggered by blanking.
What I find odd is if the screen is *not* blanked then I have to run
"dpms force off" twice. The first time *start* the latency.
Odd.
BTW -- yes the laptop is plugged in.
> You also wrote that the MHz rating as exported in /proc/cpuinfo remains
> constant, so the CPU's P state isn't changing. You may be, however,
> seeing a problem with C states. If you're kernel is so configured,
> /proc/acpi/processor/CPU0/performance will show the P states and
> /proc/acpi/processor/CPU0/power the C states. Check whether those
> change when the screen blanks.
/proc/acpi/processor/CPU0/performance: No such file or directory
> Is there any change in the rate of interupts when the latency occurs?
> Try triggering the latency while running vmstat(1) on the laptop.
I see no change in vmstat when the latency starts.
Well, maybe a little. This starts off with latency showing and then I
ran xset dpms force off about where I added the line break. Dose not
seem significant.
moseley at tiger:~$ vmstat 1
procs -----------memory---------- ---swap-- -----io---- -system-- ----cpu----
r b swpd free buff cache si so bi bo in cs us sy id wa
0 0 0 1749552 14044 176404 0 0 23 8 69 49 0 0 99 1
0 0 0 1749536 14044 176404 0 0 0 0 38 38 0 0 100 0
0 0 0 1749544 14060 176400 0 0 0 260 61 110 0 0 100 0
0 0 0 1749544 14060 176404 0 0 0 0 35 38 0 0 100 0
0 0 0 1749544 14060 176404 0 0 0 0 78 126 0 0 100 0
0 0 0 1749536 14060 176404 0 0 0 0 46 62 0 0 100 0
0 0 0 1749544 14060 176404 0 0 0 0 58 101 0 0 100 0
1 0 0 1749544 14060 176404 0 0 0 0 30 31 0 0 100 0
0 0 0 1749544 14060 176404 0 0 0 0 61 102 0 0 100 0
2 0 0 1749544 14060 176404 0 0 0 0 35 37 0 0 100 0
0 0 0 1749620 14068 176396 0 0 0 16 151 369 1 0 99 0
0 0 0 1749660 14068 176404 0 0 0 0 125 35 0 0 100 0
0 0 0 1749660 14068 176404 0 0 0 0 157 103 0 0 100 0
0 0 0 1749668 14068 176404 0 0 0 0 128 41 0 0 100 0
0 0 0 1749668 14068 176404 0 0 0 0 156 97 0 0 100 0
0 0 0 1749668 14068 176404 0 0 0 0 131 38 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 196 158 114 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 0 132 39 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 0 157 107 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 0 138 39 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 0 159 99 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 0 130 38 0 0 100 0
1 0 0 1749668 14084 176404 0 0 0 0 158 99 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 0 130 42 0 0 100 0
0 0 0 1749668 14084 176404 0 0 0 0 165 111 0 0 100 0
0 0 0 1749660 14084 176404 0 0 0 0 135 37 0 0 100 0
--
Bill Moseley
moseley at hank.org
More information about the xorg
mailing list