Date: Sat, 3 Jan 2004 03:55:42 -0800 From: Alfred Perlstein <alfred@freebsd.org> To: "David G. Lawrence" <dg@root.com> Cc: hackers@freebsd.org Subject: Re: [PATCH] sendfile erroniously returns ENOTCONN. Message-ID: <20040103115542.GW9623@elvis.mu.org> In-Reply-To: <20040103085515.GR213@nexus.dglawrence.com> References: <20040103005338.GU9623@elvis.mu.org> <20040103054115.GV56722@nexus.dglawrence.com> <20040103060156.GV9623@elvis.mu.org> <20040103085515.GR213@nexus.dglawrence.com>
next in thread | previous in thread | raw e-mail | index | archive | help
* David G. Lawrence <dg@root.com> [040103 00:55] wrote: > > * David G. Lawrence <dg@dglawrence.com> [040102 21:41] wrote: > > > > > > sendfile(8) tries to maintain compatibility with sosend as much as is > > > reasonable. ENOTCONN is the appropriate error to return if the socket > > > isn't connected. sosend checks SS_CANTSENDMORE prior to the check for > > > SS_ISCONNECTED, however, and returns EPIPE in that case. Perhaps sendfile > > > should be changed to do the same (just a though - I'm not proposing > > > that this be done). > > > Removing the check entirely seems clearly wrong, however. > > > > I had forgotten that sendfile bypasses sosend(9). I could > > add the check, is there a reason not to? The one reason I > > figured was that sometimes blocking sigpipe can be hairy inside > > libraries. Now that we can selectively disable SIGPIPE using > > the setsockopt using Apple's code this is less of an issue. > > Yes, I think checking for SS_CATSENDMORE (and returning EPIPE) prior to > checking SS_ISCONNECTED (and returning ENOTCONN as it does now) is the right > thing to do. Last question (I hope)... :) Why not call sosend? -- - Alfred Perlstein - Research Engineering Development Inc. - email: bright@mu.org cell: 408-480-4684
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040103115542.GW9623>