Can someone please steer me to the appropriate readings -- either in some doc files somewhere or into the appropriate code? (I hope that someone has written a detailed explanation that answers my question without the need to read and desk-execute pages of arcane 'C' or similar.)

My program uses fopen( "/myPath/myFile.dat" ) or one of the dozens of related calls that asks for access to a file on disk. In the case of a literal call to "fopen( )" it lives in the 'C' run-time-library.

Somewhere in that library, there are calls to more detailed services.
In no particular order:

• which file system is involved
• do we need other services, like auto-mount, to make the
  requested file system available
• how does that file system handle file paths
• how does that file system handle file names
• how does that file system handle ownership
• how does that file system handle permissions
• are access control lists used in addition to "permissions"

The services perform a pachinko-machine-like set of decisions that either grant or deny my request to access the named file. If my request gets granted, then the services pass some sort of "file handle" or "file pointer" or "file object" back to my code.

Thanks in advance,
~~~ 8d;-/ Dan

The C library ends up calling the open() system call in the kernel which is where most of what you describe takes place.

1 members found this post helpful.

Thanks for confirming what I suspected. How do I discover where to find this system call? Okay, grep -e "open\(*\)" *.c or similar...

Seriously, is there any kernel code reading support available as developer workbench or tools? Can someone get me close in terms of "components" or "sub-systems" or something?

Is there some sort of kernel code "road map" or "index" to system calls?

Thanks, again,
~~~ 0;-Dan

Relevant reading http://www.linuxtopia.org/online_boo...nel_index.html