Messages in this thread |  | | Date | Mon, 23 Oct 2000 12:52:56 -0700 | From | "David S. Miller" <> | Subject | Re: syslog() blocks on glibc 2.1.3 with kernel 2.2.x |
| |
From: patl@cag.lcs.mit.edu (Patrick J. LoPresti) Date: 23 Oct 2000 15:58:39 -0400
Is the Linux behavior even consistent with POSIX? You are saying that SOCK_DGRAM behaves fundamentally differently for local connections than for remote ones; in particular, the latter *do not block* while the former do. Even if this is standard, it ought to be documented better...
I did not say whether such calls block or not, during memory allocation they might.
Look, show me how memcpy() can drop a packet and then we can continue on about whether SOCK_DGRAM on AF_UNIX should behave one way or the other ok? :-) Because this is all that AF_UNIX is, memcpy, like a pipe. Therefore, the reliability distincion between SOCK_DGRAM/SOCK_STREAM has no meaning for AF_UNIX sockets.
Later, David S. Miller davem@redhat.com - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
|  |