View previous topic :: View next topic |
Author |
Message |
zwergzonk n00b
Joined: 06 Apr 2005 Posts: 5
|
Posted: Wed Apr 06, 2005 2:41 pm Post subject: |
|
|
That's exactly what I wanted, thnx |
|
Back to top |
|
|
zwergzonk n00b
Joined: 06 Apr 2005 Posts: 5
|
Posted: Thu Apr 07, 2005 12:13 am Post subject: |
|
|
I applied the patch set Polling Rate to 2, compiled, installed and bootet the new Kernel.
cat /proc/bus/usb/devices Output: Code: |
T: Bus=02 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 3 Spd=1.5 MxCh= 0
D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs= 1
P: Vendor=1532 ProdID=0001 Rev= 1.00
S: Manufacturer=Razer
S: Product=Razer Diamonback Optical Mouse
C:* #Ifs= 1 Cfg#= 1 Atr=a0 MxPwr=100mA
I: If#= 0 Alt= 0 #EPs= 1 Cls=03(HID ) Sub=01 Prot=02 Driver=usbhid
E: Ad=81(I) Atr=03(Int.) MxPS= 8 Ivl=10ms |
However cat /sys/module/usbhid/parameters/mousepoll: |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Thu Apr 07, 2005 6:35 am Post subject: |
|
|
zwergzonk wrote: | I applied the patch set Polling Rate to 2, compiled, installed and bootet the new Kernel.
cat /proc/bus/usb/devices Output: Code: |
T: Bus=02 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 3 Spd=1.5 MxCh= 0
D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS= 8 #Cfgs= 1
P: Vendor=1532 ProdID=0001 Rev= 1.00
S: Manufacturer=Razer
S: Product=Razer Diamonback Optical Mouse
C:* #Ifs= 1 Cfg#= 1 Atr=a0 MxPwr=100mA
I: If#= 0 Alt= 0 #EPs= 1 Cls=03(HID ) Sub=01 Prot=02 Driver=usbhid
E: Ad=81(I) Atr=03(Int.) MxPS= 8 Ivl=10ms |
However cat /sys/module/usbhid/parameters/mousepoll: |
Yeah, sorry, that is the expected behavior. The information you can see in usbfs, is the device descriptor itself, and this is seen as read-only. Yes, I know earlier versions did the interval-chaning directly in them, but it was somewhat unclean.
Excuse the confusion. |
|
Back to top |
|
|
zwergzonk n00b
Joined: 06 Apr 2005 Posts: 5
|
Posted: Thu Apr 07, 2005 1:10 pm Post subject: |
|
|
So it did work? |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Thu Apr 07, 2005 3:06 pm Post subject: |
|
|
zwergzonk wrote: | So it did work? |
Yes. |
|
Back to top |
|
|
juazp n00b
Joined: 09 Mar 2005 Posts: 6
|
Posted: Fri Apr 15, 2005 6:00 am Post subject: |
|
|
doesn't seem to apply on 2.6.11-r6 on amd64 nor did on r4 , gento-sources that is .
looks like it's built in actually -.- , time to do some testing |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Fri Apr 15, 2005 12:20 pm Post subject: |
|
|
juazp wrote: | doesn't seem to apply on 2.6.11-r6 on amd64 nor did on r4 , gento-sources that is .
looks like it's built in actually -.- , time to do some testing |
No, it should not be built into 2.6.11-kernels, other than those greater than 2.6.11-bk7.
If you want to avoid patching, I recommend you to use 2.6.12-rc2, which includes the functionality by default.
But, back to your question. Which patch were you exactly trying to apply? |
|
Back to top |
|
|
jordeh n00b
Joined: 31 May 2005 Posts: 1
|
Posted: Tue May 31, 2005 4:43 pm Post subject: |
|
|
Can some1 tell me how to install this i dont have a clue
Cheers |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Thu Jun 02, 2005 12:09 pm Post subject: |
|
|
jordeh wrote: | Can some1 tell me how to install this i dont have a clue
Cheers |
Hmm, right now, I think it's best to recommend you to use 2.6.12 when it's released. (You could also just basically use any of the 2.6.12-rc kernels if you'd like.)
You don't really have to "install" anything. What you do, is that you pass either a boot parameter to your kernel, or a module parameter to the usbhid module, all depending on whether you built USB HID support into the kernel, or as a module.
I'd have to recommend you to look inside the Documentation/kernel-parameters.txt file, for more information on this. This specific parameter should be listed with the text "[USBHID]" next to it, so use that as a search-string. Also, the uppermost part of kernel-parameters.txt should explain you how to use the parameter, or how to extract exactly the one you're going to need.
If still are confused, then please, post again! |
|
Back to top |
|
|
gaminggeek Apprentice
Joined: 13 Nov 2003 Posts: 231 Location: New Zealand
|
Posted: Fri Jun 03, 2005 5:17 am Post subject: |
|
|
Ok I got a nice new wireless mouse
but will this overclocking work on it and will it eat my batterys cause I still wanna keep my 6 moth batter time.Oceania |
|
Back to top |
|
|
Arataki n00b
Joined: 30 Oct 2004 Posts: 9 Location: South Africa
|
Posted: Thu Jun 16, 2005 12:11 am Post subject: |
|
|
its pointless using the polling rate hack on a wireless mouse as they are limited to about a 150mhz polling rate because of the wireless link. |
|
Back to top |
|
|
gaminggeek Apprentice
Joined: 13 Nov 2003 Posts: 231 Location: New Zealand
|
Posted: Thu Jun 16, 2005 8:29 am Post subject: |
|
|
oh |
|
Back to top |
|
|
tosk n00b
Joined: 06 Jun 2005 Posts: 18
|
Posted: Mon Jun 20, 2005 6:14 am Post subject: |
|
|
Arataki wrote: | its pointless using the polling rate hack on a wireless mouse as they are limited to about a 150mhz polling rate because of the wireless link. |
I actually HAD to use this hack to get my wireless optical mouse to work properly. Before it was erratic when moving. I assume this is because the polling rate was too low. I set the rate to 2ms and now it's all fluid motion.
Just for those who are complete newbs (which some are ) or for those who may not know exactly what to do:
For builtin USB HID, pass this as a kernel parameter in grub or lilo:
I'm not sure about HID as a module. I've never actually passed arguments to loadable modules before.
edit: is it technically a hack anymore since it's part of the kernel? _________________ Tosk |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Mon Jun 20, 2005 11:23 am Post subject: |
|
|
tosk wrote: | Arataki wrote: | its pointless using the polling rate hack on a wireless mouse as they are limited to about a 150mhz polling rate because of the wireless link. |
I actually HAD to use this hack to get my wireless optical mouse to work properly. Before it was erratic when moving. I assume this is because the polling rate was too low. I set the rate to 2ms and now it's all fluid motion.
Just for those who are complete newbs (which some are ) or for those who may not know exactly what to do:
For builtin USB HID, pass this as a kernel parameter in grub or lilo:
I'm not sure about HID as a module. I've never actually passed arguments to loadable modules before.
|
TFM (still) says:
file:Documentation/kernel-parameters.txt wrote: | Module parameters for loadable modules are specified only as the
parameter name with optional '=' and value as appropriate, such as:
modprobe usbcore blinkenlights=1
Module parameters for modules that are built into the kernel image
are specified on the kernel command line with the module name plus
'.' plus parameter name, with '=' and value if appropriate, such as:
usbcore.blinkenlights=1 |
So the module parameter is: Code: | modprobe usbhid mousepoll=<foo> |
Oh, and while I'm at it, I should probably also mention that people should watch out how small of an interval they're using. Well, no matter what, the (really) only available intervals are:
(There are, of course, greater ones, but it would be stupid to decrease the preciseness of ones mouse)
8 = 125Hz,
4 = 250Hz,
2 = 500Hz,
1 = 1KHz.
If other intervals are passed, they'll just rounded down. Also, I suggest that people play with different intervals. I find myself using 4 nowadays, for two simple reasons:
1) I can't feel the difference.
2) 250 interrupts per second is a lot less than 500 or 1000, but still not amazingly more than 125.
So, think about how many IRQs your mouse is able to fire off at any given interval before you pick it. It might not be worth it having both the timer interrupt running thousand times per second, *AND* have your mouse do it too.
tosk wrote: | edit: is it technically a hack anymore since it's part of the kernel? |
Of course it is. The kernel is just one big hack. |
|
Back to top |
|
|
Arataki n00b
Joined: 30 Oct 2004 Posts: 9 Location: South Africa
|
Posted: Sat Jun 25, 2005 5:15 pm Post subject: |
|
|
Tosk: you probably noticed a difference because the 150MHz of the wireless wasn't synced to the 125MHz polling rate of the usb. You could probably drop it to 4ms and it would feel the same.
But I may be wrong. I'll have to find that article I read detailing all the polling rate stuff. |
|
Back to top |
|
|
Cinder6 l33t
Joined: 05 Aug 2004 Posts: 767 Location: California
|
Posted: Thu Jul 07, 2005 3:12 am Post subject: |
|
|
for what it's worth, here is the info for the mx518: Code: | { VENDOR_LOGITECH, 0xC01E, "M-BS81A", "MX518 Optical Mouse", HAS_RES | HAS_SS | HAS_SSR }, |
Thought it was odd it has the same M/N as the mx510, but that's what it says on the mouse.
Note: this will not enable cruise control for some reason. You still get the res change if you use those buttons. If anyone has any ideas... _________________ Knowledge is power.
Power corrupts.
Study hard.
Be evil.
Ugly Overload |
|
Back to top |
|
|
mli n00b
Joined: 24 Jul 2004 Posts: 18 Location: Finland
|
Posted: Thu Jul 07, 2005 9:02 pm Post subject: |
|
|
I just emerged gentoo-sources-2.6.12-r4 and I have usbhid.mousepoll=2 added to kernel parameters. cat /sys/module/usbhid/parameters/mousepoll shows that mousepoll is set to 2, but cat /proc/bus/usb/devices shows that my mouse is using 10ms interval. Usbview shows 10ms intervall too. What might be wrong? I'm using logitechs MX300. |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Thu Jul 07, 2005 9:29 pm Post subject: |
|
|
mli wrote: | I just emerged gentoo-sources-2.6.12-r4 and I have usbhid.mousepoll=2 added to kernel parameters. cat /sys/module/usbhid/parameters/mousepoll shows that mousepoll is set to 2, but cat /proc/bus/usb/devices shows that my mouse is using 10ms interval. Usbview shows 10ms intervall too. What might be wrong? I'm using logitechs MX300. |
This is the expected behavior. /proc/bus/usb/devices shows the device descriptor (the default values) hard-coded into the device.
It's true that old versions of the patch changed the interval directly in the descriptor (and thus changed what could be seen in /proc/bus/usb/devices), but this been changed, since it wasn't a Good Thing (tm).
So, yes, everything's fine. |
|
Back to top |
|
|
azuriel Apprentice
Joined: 27 Feb 2005 Posts: 166
|
Posted: Sun Jul 17, 2005 5:45 pm Post subject: |
|
|
This is quoted from an article at Gamespot about all this...I have a Logitech MX518. Quote: | Recently, users have taken to hacking their USB ports to increase the polling rate of USB ports to 250Hz and even 500Hz in an attempt to rid themselves of negative acceleration. This technique does result in some success but the faster polling speeds also sacrifice system compatibility with many other USB devices. According to Logitech, increasing USB frequencies can tax the CPU an extra 5 to 10 percent.
Instead of eating up precious CPU cycles and destroying compatibility with other USB devices, Logitech opted to widen the data path. By using a 16-bit data path and the standard 125Hz transmission rate, Logitech's MX 518 Gaming-Grade Mouse can transmit a change in position of 32,000 pixels at each polling interval, solving the problem of negative acceleration without increasing the polling rate. |
From what the article says, making the polling rate faster won't give me much of an effect. The dpi changer in the mouse is hardware, so I can go from 400 to 800 to 1600 dpi just fine. And it seems that Logitech tried to make a product where the effect of changing the polling rate is negligible.
Edit: Also found this line in another article. Quote: | PS/2 users will be able to benefit by increasing their polling rate to a max of 200 polls/sec which is the max allowed in PS/2. The side effect of increasing the polling rate in PS/2 mode is that it taxes your cpu significantly so most gamers will want to stick to USB's 125 polls/sec. |
So far I really see no reason to mess with this setting. _________________ Adopt an unanswered post
TJGames.org
The folly of mistaking a torrent of verbiage for a spring of capital truths, and oneself for an oracle, is inborn in us. -Valery |
|
Back to top |
|
|
m0p Apprentice
Joined: 20 Jun 2005 Posts: 205 Location: en_GB
|
Posted: Sun Jul 17, 2005 7:17 pm Post subject: |
|
|
I've enabled USB polling as 2ms in the kernel, but usbview says the interval is 10ms. Got any ideas? |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Sun Jul 17, 2005 7:25 pm Post subject: |
|
|
azuriel wrote: | This is quoted from an article at Gamespot about all this...I have a Logitech MX518. Quote: | Recently, users have taken to hacking their USB ports to increase the polling rate of USB ports to 250Hz and even 500Hz in an attempt to rid themselves of negative acceleration. This technique does result in some success but the faster polling speeds also sacrifice system compatibility with many other USB devices. According to Logitech, increasing USB frequencies can tax the CPU an extra 5 to 10 percent.
Instead of eating up precious CPU cycles and destroying compatibility with other USB devices, Logitech opted to widen the data path. By using a 16-bit data path and the standard 125Hz transmission rate, Logitech's MX 518 Gaming-Grade Mouse can transmit a change in position of 32,000 pixels at each polling interval, solving the problem of negative acceleration without increasing the polling rate. |
From what the article says, making the polling rate faster won't give me much of an effect. The dpi changer in the mouse is hardware, so I can go from 400 to 800 to 1600 dpi just fine. And it seems that Logitech tried to make a product where the effect of changing the polling rate is negligible.
Edit: Also found this line in another article. Quote: | PS/2 users will be able to benefit by increasing their polling rate to a max of 200 polls/sec which is the max allowed in PS/2. The side effect of increasing the polling rate in PS/2 mode is that it taxes your cpu significantly so most gamers will want to stick to USB's 125 polls/sec. |
So far I really see no reason to mess with this setting. |
Who are you trying to convince? Yourself?
I really can't recommend anything but trying it out, before ruling it out. |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Sun Jul 17, 2005 7:25 pm Post subject: |
|
|
m0p wrote: | I've enabled USB polling as 2ms in the kernel, but usbview says the interval is 10ms. Got any ideas? |
Please see some of the other posts on this page. (Hint: It's expected.) |
|
Back to top |
|
|
azuriel Apprentice
Joined: 27 Feb 2005 Posts: 166
|
Posted: Sun Jul 17, 2005 7:30 pm Post subject: |
|
|
I guess I will, next time I feel obliged to recompile my kernel. Anything to maximize those headshots, eh? I'm just trying not to set unrealistic expectations for improvement. _________________ Adopt an unanswered post
TJGames.org
The folly of mistaking a torrent of verbiage for a spring of capital truths, and oneself for an oracle, is inborn in us. -Valery |
|
Back to top |
|
|
m0p Apprentice
Joined: 20 Jun 2005 Posts: 205 Location: en_GB
|
Posted: Sun Jul 17, 2005 7:30 pm Post subject: |
|
|
Ohhh, so I should look in /sys/module/usbhid/parameters/mousepoll for the real polling rate? |
|
Back to top |
|
|
krejler Tux's lil' helper
Joined: 10 Nov 2003 Posts: 142 Location: Denmark
|
Posted: Sun Jul 17, 2005 7:40 pm Post subject: |
|
|
m0p wrote: | Ohhh, so I should look in /sys/module/usbhid/parameters/mousepoll for the real polling rate? |
Well, that will just give you the value *you* specified. Right now, there isn't really a way to query to mouse to get the value *it* is using.
The best way to test if the "hack" worked, is to simlpy move your mouse around a bit, and try to feel it. |
|
Back to top |
|
|
|