Date: Sat, 20 Dec 2008 18:49:21 -0800 From: Sean Bruno <sbruno@miralink.com> To: Dieter <freebsd@sopwith.solgatos.com> Cc: freebsd-firewire@freebsd.org, bug-followup@freebsd.org Subject: Re: kern/118093: firewire bus reset hogs CPU, causing data to be lost Message-ID: <494DAEB1.1070909@miralink.com> In-Reply-To: <200812191902.TAA01939@sopwith.solgatos.com> References: <200812191902.TAA01939@sopwith.solgatos.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Dieter wrote:
> Here are the results of two more experiments:
>
> I tried to verify what spl the firewire code is at when calling
> printf(9) by adding the following block into fwohci.c just
> before one of the printfs.
>
> {
> #include <sys/types.h>
> #include <sys/systm.h>
> intrmask_t debug_spl; /* __uint32_t */
> intrmask_t debug_spl_high;
> intrmask_t debug_spl_tty;
> intrmask_t debug_spl_fw;
> intrmask_t debug_spl_0;
> intrmask_t debug_spl_net;
>
> debug_spl = splhigh();
> debug_spl_high = spltty();
> debug_spl_tty = splfw();
> debug_spl_fw = splnet();
> debug_spl_net = splhigh();
> spl0(); /* void */
> debug_spl_0 = splhigh();
> splx(debug_spl);
>
> printf("fwohci_intr_core(): spl = 0x%x\n splhigh=0x%x spltty=0x%x splfw=0x%x splnet=0x%x spl0=0x%x\n",
> debug_spl, debug_spl_high, debug_spl_tty, debug_spl_fw, debug_spl_net, debug_spl_0);
> }
>
> But my results appear bogus:
>
> fwohci_intr_core(): spl = 0x0
> splhigh=0x0 spltty=0x0 splfw=0x0 splnet=0x0 spl0=0x0
>
> I have examined my code and the spl(9) man page several times
> but I can't find what is wrong. Any clues?
>
> --------------------------------------------------------
>
> To isolate the effects of printf(9) from the firewire bus reset,
> I picked a trivial system call (chown(2)) and added some printf(9)
> calls.
>
> Calling chown several times and monitoring with systat -vmstat gives:
> 1098 interrupts on the console IRQ
> 93.1%Sys 6.7%Intr 0.2%User 0.0%Nice 0.0%Idle
>
> This did NOT interfere with Ethernet.
>
> So printf(9) interferes with Ethernet when called from the
> firewire driver, but not when called from a vanilla system call.
>
>
I setup my system to execute a bus reset every 1 second, simultaneously,
I started copying a large
file onto the system see if anything would happen.
I saw no change to copy speed reported by SCP during the entire
transaction. I also see no change
to the system load while this is occurring.
This seems indicative of "something else" going on.
> --------------------------------------------------------
>
> sio0: <16550A-compatible COM port> port 0x3f8-0x3ff irq 4 flags 0x10 on acpi0
> sio0: type 16550A, console
> sio0: [FILTER]
>
> What does [FILTER] mean? I don't see an explanation on the sio man page.
>
Did you mean to ask a different list? Because I have no idea. :)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?494DAEB1.1070909>