I did a clean build, and now I'm not seeing the problem, so it must have been something I messed up in my build. Not sure what I did to cause it; so, this patch should be discarded. Interesting to me though that the "dma_sync_single_for_cpu" is generally doing nothing on MIPS.
Nathan On Tue, 2013-02-12 at 03:10 +0100, Andreas Mohr wrote: > Hi, > > On Mon, Feb 11, 2013 at 12:00:01PM +0100, > [email protected] wrote: > > Date: Sun, 10 Feb 2013 22:40:12 -0800 > > From: Nathan Hintz <[email protected]> > > > > +-static inline int cpu_is_noncoherent_r10000(struct device *dev) > > ++static inline int cpu_is_noncoherent(struct device *dev) > > + { > > + return !plat_device_is_coherent(dev) && > > + (current_cpu_type() == CPU_R10000 || > > +- current_cpu_type() == CPU_R12000); > > ++ current_cpu_type() == CPU_R12000 || > > ++ current_cpu_type() == CPU_74K); > > + } > > Would this also (not yet??) apply to > > # cat /proc/cpuinfo > system type : Broadcom BCM47XX > processor : 0 > cpu model : Broadcom BCM3302 V2.9 > BogoMIPS : 239.20 > wait instruction : yes > microsecond timers : yes > tlb_entries : 32 > extra interrupt vector : yes > hardware watchpoint : no > ASEs implemented : mips16 > shadow register sets : 1 > core : 0 > VCED exceptions : not available > VCEI exceptions : not available > > > Linux version 2.6.34.5 (andi@note) (gcc version 4.3.3 (GCC) ) #2 Fri Sep > 17 01:28:35 CEST 2010 > CPU revision is: 00029029 (Broadcom BCM3302) > > > (ASUS WL-500gP V2; yeah, that installation is ooold...) > > This system is noncoherent, right? (judging from earlier discussions, > but perhaps that's wrong) > If it is, then it would have to be marked, but I just cannot really > figure out the actual cputype setting from the myriads of cputype > assignments in cpu-probe.c, to see whether cpu_is_noncoherent() already > includes the cputype which this system ends up with... > At least it's PRID_COMP_BROADCOM, judging from CPU revision 00029029. > > > It might also be a useful idea to add some per-cputype comments to this > helper why (or why not) a certain CPU is thought to be (non-)coherent > (with a most authoritative reference added each if available, ideally!). > > Thanks, > > Andreas Mohr > _______________________________________________ openwrt-devel mailing list [email protected] https://lists.openwrt.org/mailman/listinfo/openwrt-devel
