Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

conmon: drop usage of splice(2) #131

Merged
merged 3 commits into from
Mar 20, 2020
Merged

conmon: drop usage of splice(2) #131

merged 3 commits into from
Mar 20, 2020

Commits on Mar 18, 2020

  1. .dir-locals.el: new file 

    Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
    @giuseppe
    giuseppe committed Mar 18, 2020
    Configuration menu
    Copy the full SHA
    a2e1060 View commit details
    Browse the repository at this point in the history

  2. conmon: drop usage of splice(2) 

    splice(2) doesn't seem to handle correctly sockets having a type of
SOCK_SEQPACKET.

When using a SOCK_SEQPACKET, each read is limited to the current
packet.  The data that doesn't fit in a single read(2) call will be
discarded.  This can be a problem when the fd_out doesn't accept all
the data, so on a short write we lose the data that could not be
written by the splice(2) call.

Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
    @giuseppe
    giuseppe committed Mar 18, 2020
    Configuration menu
    Copy the full SHA
    6da0481 View commit details
    Browse the repository at this point in the history

  3. conmon: avoid hanging on stdin 

    now that we dropped splice(2), we need to make sure the read/write
loop doesn't hang indefinitely conmon.

Attempting to write to a fd that is not ready for write causes conmon
to block on the write(2) call, at the same time the container process
might be blocked attempting to write to stderr but since conmon is not
processing any data from the container, both processes are dead
locked.

Keep in a buffer what we read from the pipe and write to the
masterfd_stdin only when it is ready for write.

Signed-off-by: Giuseppe Scrivano <gscrivan@redhat.com>
    @giuseppe
    giuseppe committed Mar 18, 2020
    Configuration menu
    Copy the full SHA
    6cf14cb View commit details
    Browse the repository at this point in the history