Date: Wed, 14 Jul 2004 17:01:20 +0400 From: Gleb Smirnoff <glebius@freebsd.org> To: Robert Watson <rwatson@freebsd.org> Cc: current@freebsd.org Subject: Re: Some netgraph node global locking patches Message-ID: <20040714130120.GA7897@cell.sick.ru> In-Reply-To: <Pine.NEB.3.96L.1040714002717.83353E-100000@fledge.watson.org> References: <Pine.NEB.3.96L.1040714002717.83353E-100000@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Jul 14, 2004 at 12:30:40AM -0400, Robert Watson wrote: R> //depot/vendor/freebsd/src/sys/netgraph/ng_eiface.c R> //depot/vendor/freebsd/src/sys/netgraph/ng_fec.c R> //depot/vendor/freebsd/src/sys/netgraph/ng_iface.c Well, these three are quite straightforward and identical. Look fine. R> //depot/vendor/freebsd/src/sys/netgraph/ng_ppp.c Is there any hidden obstacles for merging qsort_r() from libc to libkern? It will help to remove this ugly hack. R> //depot/vendor/freebsd/src/sys/netgraph/ng_pppoe.c Agreed with comment. Anyway, I'm planning to move this configuration trigger to private data. sysctl's are not very elegant way to set configaration of netgraph node. Moreover, I can imagine setup when you need to serve non-standard PPPoE only on one interface, and normal PPPoE on other interfaces: for example two different networks merged to use one AC. R> //depot/vendor/freebsd/src/sys/netgraph/ng_tty.c R> +/* R> + * XXXRW: ngt_unit is protected by ng_tty_mtx. ngt_ldisc is constant once R> + * ng_tty is initialized. ngt_nodeop_ok is untouched, and might want to be a R> + * sleep lock in the future? R> + */ One question: are any locks held when linesw callbacks (ngt_open, ngt_close, etc..) are called? -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPE
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040714130120.GA7897>