From 7c49988ebf5c176cadd4a9e287e443d49a2cdeec Mon Sep 17 00:00:00 2001 From: Eric Wong Date: Wed, 17 Jul 2013 01:39:24 +0000 Subject: document ioq and mog_{mgmt,http}_drop interaction safety I needed to spend time to convince myself this was safe, so leave a note to others (and future self) in case there is cause for concern. Basically, this is highly dependent on our overall one-shot-based concurrency model and safe as long as basic rules are followed. --- http.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) (limited to 'http.c') diff --git a/http.c b/http.c index 19cf9b7..cc802c3 100644 --- a/http.c +++ b/http.c @@ -122,7 +122,14 @@ void mog_http_unlink_ftmp(struct mog_http *http) file->tmppath, http->svc->docroot); } -/* called if epoll/kevent is out-of-space */ +/* + * called only if epoll/kevent is out-of-space + * Note: it's impossible for this to be called while this mfd is + * inside an ioq SIMPLEQ, however mfd->ioq_blocked /may/ be true when + * this function is called. We could add validation code to ensure + * this remains true, but that would increase the size of "struct mog_fd", + * so we will rely on this comment instead. + */ void mog_http_drop(struct mog_fd *mfd) { struct mog_http *http = &mfd->as.http; -- cgit v1.2.3-24-ge0c7