Date: Thu, 23 May 2013 16:05:39 -0500 From: Guy Helmer <guy.helmer@gmail.com> To: John Baldwin <jhb@freebsd.org> Cc: freebsd-net@freebsd.org Subject: Re: bpf hold buffer in-use flag Message-ID: <4EA47178-7CE2-40CE-A767-2689FAF7BEBD@gmail.com> In-Reply-To: <201301091535.04904.jhb@freebsd.org> References: <9C928117-2230-4F01-9B95-B6D945AF4416@gmail.com> <201301091535.04904.jhb@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Jan 9, 2013, at 2:35 PM, John Baldwin <jhb@freebsd.org> wrote: > On Tuesday, November 13, 2012 4:40:57 pm Guy Helmer wrote: >> To try to completely resolve the race in bpfread(), I have put = together=20 > these changes to add a flag to indicate when the hold buffer cannot be=20= > modified because it is in use. Since it's my first time using = mtx_sleep() and=20 > wakeup(), I wanted to run these past the list to see if I can get any = feedback=20 > on the approach. >>=20 >>=20 >> Index: bpf.c >> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D >> --- bpf.c (revision 242997) >> +++ bpf.c (working copy) >> @@ -819,6 +819,7 @@ bpfopen(struct cdev *dev, int flags, int fmt, = stru >> * particular buffer method. >> */ >> bpf_buffer_init(d); >> + d->bd_hbuf_in_use =3D 0; >> d->bd_bufmode =3D BPF_BUFMODE_BUFFER; >> d->bd_sig =3D SIGIO; >> d->bd_direction =3D BPF_D_INOUT; >> @@ -872,6 +873,9 @@ bpfread(struct cdev *dev, struct uio *uio, int = iof >> callout_stop(&d->bd_callout); >> timed_out =3D (d->bd_state =3D=3D BPF_TIMED_OUT); >> d->bd_state =3D BPF_IDLE; >> + while (d->bd_hbuf_in_use) >> + mtx_sleep(&d->bd_hbuf_in_use, &d->bd_lock, >> + PRINET|PCATCH, "bd_hbuf", 0); >=20 > You need to check the return value here, otherwise the PCATCH is = useless (you=20 > will just go back to sleep instead of failing with an error if this is=20= > interrupted by a signal).=20 Thanks for the feedback (sorry it's taken so long to get to it). Would = this change correctly handle interruptions? Index: bpf.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- bpf.c (revision 250941) +++ bpf.c (working copy) @@ -856,9 +856,14 @@ callout_stop(&d->bd_callout); timed_out =3D (d->bd_state =3D=3D BPF_TIMED_OUT); d->bd_state =3D BPF_IDLE; - while (d->bd_hbuf_in_use) - mtx_sleep(&d->bd_hbuf_in_use, &d->bd_lock, + while (d->bd_hbuf_in_use) { + error =3D mtx_sleep(&d->bd_hbuf_in_use, &d->bd_lock, PRINET|PCATCH, "bd_hbuf", 0); + if (error =3D=3D EINTR || error =3D=3D ERESTART) { + BPFD_UNLOCK(d); + return (error); + } + } /* * If the hold buffer is empty, then do a timed sleep, which * ends when the timeout expires or when enough packets
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4EA47178-7CE2-40CE-A767-2689FAF7BEBD>