Date: Thu, 23 May 2002 13:37:26 -0700 From: Conrad Minshall <conrad@apple.com> To: "Semen A. Ustimenko" <semenu@FreeBSD.ORG>, freebsd-net@FreeBSD.ORG Cc: freebsd-fs@FreeBSD.ORG Subject: Re: NFS don't set sopt.sopt_dir sometimes... Maybe sosetopt() should? Message-ID: <l03130304b912feb55ad8@[17.219.180.26]> In-Reply-To: <20020523111658.N405-100000@def.the.net>
next in thread | previous in thread | raw e-mail | index | archive | help
At 9:28 PM -0700 5/22/02, Semen A. Ustimenko wrote: >Looks like nfs_socket.c and nfs_syscalls.c lack strings > > sopt.sopt_dir = SOPT_SET; > >when setting TCP_NODELAY and SO_KEEPALIVE. For SO_KEEPALIVE, it doesn't >matter, sosetopt() doesn't examine it, but TCP_NODELAY is actually >ignored. > >Obviously, it's easy to add these lines, but maybe it's better to make >sosetopt() set sopt_dir for callers? This came up with SMB (and NFS) in Darwin. Changing both caller and callee is the robust choice in our case as old loadable kernel module binaries with new kernels will get fixed, and vice-versa. If you have the luxury of not considering out-of-sync kernel loadables then I envy you :) -- Conrad Minshall, conrad@apple.com, 408 974-2749 Apple Computer, Mac OS X Core Operating Systems To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?l03130304b912feb55ad8>