Follow-up Comment #1, bug #26960 (project hurd):

Someone should write test cases for whether the patch causes problems with the
following flags.

* O_APPEND: Open a firmlink with O_APPEND and then use io_get_openmodes and
verify that the O_APPEND bit is there.
* O_NOATIME: Open a firmlink with O_NOATIME, use io_read, close the file_t and
check that the atime of the target file remains unchanged.
* others?

If the firmlink translator doesn't pass O_APPEND to the dir_lookup call, but
its caller ends up doing another dir_lookup and setting O_APPEND there, then
it's okay, provided that the RPC responses from the firmlink translator are
such that the caller cannot expect the file_t to work right if it omits the
second dir_lookup.


Reply to this item at:


  Message sent via/by Savannah

Reply via email to