Date: Thu, 26 Dec 1996 09:36:41 +0900 From: Kazutaka YOKOTA <yokota@zodiac.mech.utsunomiya-u.ac.jp> To: ccsanady@iastate.edu Cc: current@freebsd.org, yokota@zodiac.mech.utsunomiya-u.ac.jp Subject: Re: ps/2 mouse problems in 2.2-BETA.. Message-ID: <199612260036.JAA20114@zodiac.mech.utsunomiya-u.ac.jp> In-Reply-To: Your message of "Wed, 25 Dec 1996 15:35:50 CST." <9612252135.AA04491@isua1.iastate.edu> References: <9612252135.AA04491@isua1.iastate.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
>I have been experiencing some problems with mouse in this beta. What >seems to happen is when the system is hevily swapping, and there is a >lot of mouse movement, the cursor goes insane. It will zip around the >screen, click randomly and such for a while. The thing is, I know its not >just queued requests, because I see actions that I know I never performed >with the mouse. It's almost as though the buffer was filled with junk, >very odd. The most likely explanation is that the mouse and the X server have become out of sync due to a lost interrupt. Assuming you are using XFree86 and the X server is not accessing the mouse via `moused', you can switch away from X to a vty then come back to the X, and they will re-sync. This should work because the XFree86 server opens the mouse device only when it is in the foreground; it closes the mouse device when the user switches away from the X server, thus, flipping between the X server and a vty should close and re-open the device and give the X and the mouse a chance to re-sync. Unlike serial mice, it is not easy to ensure the PS/2 mouse and the driver are synchronized ;-< By default, the `psm' driver does not check the header byte of the 3-byte status data packet from the PS/2 mouse (in order to support the ALPS GlidePoint device). You may put `options PSM_CHECKSYNC' in the kernel configuration file to enable the minimal sync check feature in the `psm' driver. But, the check is minimal; you may still see the problem. When you do, click any mouse button without moving the mouse and the driver will find the header byte. The trick should work even when the mouse is accessed via `moused'. The `psm' driver in 2.2-BETA shares a common set of routines (kbdio) with the `syscons' driver to access the keyboard controller to which the PS/2 mouse port is wired. Because the routines are new, they need extensive testing. Please try the tricks above and tell me if they work. By the way, the `psm' driver in 2.1.X does not check the header byte either, and can easily go out of sync with the mouse. Did your system have the same problem under 2.1.X? If so, was the problem better or worse than now? Kazu >Anyone else had, or seen this? I figured it was due to the recent ps/2 >mouse/console problems.. is this possible? > >thanks, >chris > >btw, please cc me, im not on the lists over break..
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199612260036.JAA20114>