Patches for various software

This page contains unintegrated patches for various packages made by me. They were sent to maintainers, but I have received no comments about them (positive nor negative). Once they are applied, I move them here.

Send comments to lav@yars.free.net.

xgalaga-2.0.34.diff (1645 bytes)
patch for xgalaga-2.0.34. It fixes buffer overflow when real name is too long, makes it work well when no sound available and makes it compile on solaris. Run autoconf after applying.
transproxy-1.2-linux-2.4-iptables.diff (1855 bytes)
patch for transproxy-1.2 to work with linux-2.4/iptables.
linux-X11-sigcont.diff (414 bytes)
startx, switch back to text console, ^Z, X server is now suspended, then switch back to X11 - Alt-F7. Now console is dead. One user has really done this accidently.
This patch sends SIGCONT to the process controlling a virtual console before sending it switch-to signal.
linux-sbusfb-oops.diff (1100 bytes)
While playing with consoles on different frame buffers on twin-headed Sun, I discovered an oops - this patch fixes it.
linux-console-schedule.diff (958 bytes)
I have noticed on Sun Sparc Station 5 that large amount of scrolling on console (especially from telnet) causes mpg123 to skip. This seemed strange to me, because mpg123 consumes less than 50% CPU. It turned out that console code does not reschedule when needed. This patch solves it, but maybe not in pretty way.
linux-tcxfb-snow-2.4.2.diff (1088 bytes)
On my Sun SS5 I have tcx/24 frame buffer rev1.7. When X11 calls kernel ioctl to change cursor shape and color, "snow" appears on screen (anybody remembers those CGA adapters? :-)). It happens when kernel changes cursor color.
1. Snow appears on tcx/24bit rev1.7 when changing cursor color.
2. Moving cursor out of screen temporarily does not help.
So I made this patch which does two things:
1. check if old cursor color is the same as new one (majority of cases). I have used fb->cursor.color to store old cursor color, since I have not found where is is used. But maybe someone plans to use it differently, in this case some changes would be necessary.
2. wait for vertical retrace before changing cursor color. I have verified that this code really eliminates the snow. Don't know why that bit is called TCX_THC_MISC_CURS_RES.
Unfortunately, it seems that I'm the only one who sees the snow and cares, so I doubt this patch will be applied. If you do see the snow, please send a note about it to sparc-list@redhat.com.