Program equivalent of "mount" command

commands:

memory

Through the mount(2) syscall

Read the mount(2) manpage ("man 2 mount") for the syntax of the call.

MOUNT(2) Linux Programmer's Manual MOUNT(2)

NAME mount, umount - mount and unmount filesystems

SYNOPSIS #include

int mount(const char *source, const char *target, const char *filesystemtype, unsigned long mountflags, const void *data);

int umount(const char *target);

int umount2(const char *target, int flags);

DESCRIPTION mount() attaches the filesystem specified by source (which is often a device name, but can also be a directory name or a dummy) to the direc- tory specified by target.

Thanks, I got it working. But several things still confuse me. One is the filesystemtype. I used "msdos" and it worked. I was able to write a file to the memory stick and that file appeared normally when plugged into my WinXP system. But is "msdos" right? For example, should it handle long file names, or will it be limited to 8.3 names as in DOS? I thought there would be some sort of an "auto" selection like there is in the mount shell command.

The second question has to do with handling the case where no memory stick is plugged in. Of course the mount() call fails, and I check for that in my application and give the user appropriate feedback through an LCD screen. The the shell serial port also has some error messages, starting with "Device not ready" and continuing for 8-10 lines more of info about the error. I guess this is not a problem for me since the shell port is going to be disconnected during normal operation. So I can ignore whatever gets sent to it. But I would like to understand why this particular system call produces shell port error messages while other systems calls (like open) are content to let my program handle error returns.

Robert Scott Ypsilanti, Michigan

Hello,

You could try out "auto", which worked at least once for me using the mount program.

"msdos" is limited to 8.3 names; the newer system is called "vfat".

Does this happen using the mount program, too? If yes, it shouldn't be anything serious.

To open files that don't exist is (imho) normal behaviour and expected. To open a device which is not ready is related to kernel mode drivers (and thus important to the administrator, kernel and system stability). It is the same as if you're trying to use a floppy disk which has become very bad. Moreover, you might be tampering with the file system structure which could harm other running processes.

You could try to make the kernel talk less about things like this, they won't be logged then. But if the shell port is disconnected in normal operation, I wouldn't care. It might be more important to notice errors when things don't work as expected - and I would probably forget to make the kernel talk to me again. If you want to do it, you can find it in /proc/sys/kernel/printk .

Hope this helps, Sebastian

..just tried it. mount fails. "..no such device.."

I tried filesystemtype="vfat" and it worked fine with my USB memory stick. Thanks.

Robert Scott Ypsilanti, Michigan

This job can be done by udev automatically for you. It also could start your application in the case someone inserts a memory stick (if you want). Or send your application a signal, when the memory stick was inserted and successfully mounted. Or everything else you want (this includes it could also drive you crazy....).

JB