kremer@cs.odu.edu (Lloyd Kremer) (05/01/91)
We're having a strange problem with a device file under cu. The box is an HP500 running SysV, connected to an 8-port multiplexer, connected to a Batec buffer memory. We have a character special file, /dev/tty07, major=7, minor=1, and the corresponding entry in L.Devices. We can use cu direct 9600 to it and it works fine. If the file /dev/tty07 is rm'ed (accidentally or otherwise), and we then replace it using mknod /dev/tty07 c 7 1 , the file is recreated but it doesn't "work" anymore, i.e. data sent to it don't make it out the port. However, if we make up a new name for the file, such as mknod /dev/tty007 c 7 1 and adjust L.Devices accordingly, the new name works! What's going on? Who is "remembering" the previous file name? P.S. Rebooting the HP while the line is connected will also sometimes cause the port to stop working. -- Lloyd Kremer kremer@cs.odu.edu
kremer@snapdragon.cs.odu.edu (Lloyd Kremer) (05/01/91)
We're having a strange problem with a device file under cu. The box is an HP500 running SysV, connected to an 8-port multiplexer, connected to a Batec buffer memory. We have a character special file, /dev/tty07, major=7, minor=1, and the corresponding entry in L.Devices. We can use cu direct 9600 to it and it works fine. If the file /dev/tty07 is rm'ed (accidentally or otherwise), and we then replace it using mknod /dev/tty07 c 7 1 , the file is recreated but it doesn't "work" anymore, i.e. data sent to it don't make it out the port. However, if we make up a new name for the file, such as mknod /dev/tty007 c 7 1 and adjust L.Devices accordingly, the new name works! What's going on? Who is "remembering" the previous file name? P.S. Rebooting the HP while the line is connected will also sometimes cause the port to stop working. -- Lloyd Kremer kremer@cs.odu.edu