Date: Thu, 22 Mar 2012 07:34:48 -0700 From: Adrian Chadd <adrian@freebsd.org> To: John Baldwin <jhb@freebsd.org> Cc: freebsd-hackers@freebsd.org, Alexander Best <arundel@freebsd.org> Subject: Re: still issues with unreadable dmesg output on SMP systems Message-ID: <CAJ-Vmo=EO2vxVnG4_bjZTcbkoAhscOtnV9QdXFARe%2BSGtMgzVw@mail.gmail.com> In-Reply-To: <201203220813.13113.jhb@freebsd.org> References: <20120322101041.GA99431@freebsd.org> <201203220813.13113.jhb@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi, On 22 March 2012 05:13, John Baldwin <jhb@freebsd.org> wrote: > s/fixed/mostly worked around/ > > The problem is that the work around isn't perfect. =A0The root cause is s= till > not fixed. =A0avg@ has some ideas on better ways to handle this, but it i= s a bit > tricky to get this right since we also dont' want printfs to be delayed t= oo > much (and to fix this perfectly requires buffering printf output). I'm rapidly liking the brokenness of the IO. Why? Because it becomes _very freaking obvious_ when things are running concurrently. I've found many ath/net80211 bugs because of this. Having it as an option would be nice. :-) Adrian
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-Vmo=EO2vxVnG4_bjZTcbkoAhscOtnV9QdXFARe%2BSGtMgzVw>