Hi,
On Mon, 2009-11-09 at 17:54 +0100, Ralf Baechle wrote:
> On Tue, Nov 10, 2009 at 12:36:48AM +0800, Wu Zhangjin wrote:
>
> > > Grant Likely convinced me in Tokyo that this is the way to go. He intends
> > > to rewrite the FDT code into an easily re-usable library which he estimates
> > > to take a month or two after which I'd like to start using it on MIPS.
> > > This probably also means FDT support for PMON will eventually be needed.
> > >
> >
> > Thanks for your pointer, Will take a look at FDT asap.
> >
> > > Until then of course machtype=<whatever> is a fair solution.
> > >
> >
> > are you ready to apply it? then I will push the Cpufreq and Standby
> > support, and really hope we can get a full loongson2f support in the
> > mainline's 33 version ;)
>
> I had already taken the previous version into the -queue tree.
>
> What are the changes since -v1? I've done a few changes myself, mostly
> tweaking the English language bits of the patch, so incremental patches
> would be ideal.
>
Very few of changes, did you apply the changes in this one?
[PATCH v2 6/7] [loongson] lemote-2f: add reset support
about this part in arch/mips/loongson/lemote-2f/reset.c, I have removed
the array, and used the "switch...case..." instead.
+void mach_prepare_reboot(void)
+{
+ switch (mips_machtype) {
+ case MACH_LEMOTE_FL2F:
+ fl2f_reboot();
+ break;
+ case MACH_LEMOTE_ML2F7:
+ ml2f_reboot();
+ break;
+ case MACH_LEMOTE_YL2F89:
+ yl2f89_reboot();
+ break;
+ default:
+ break;
+ }
+}
+
+void mach_prepare_shutdown(void)
+{
+ switch (mips_machtype) {
+ case MACH_LEMOTE_FL2F:
+ fl2f_shutdown();
+ break;
+ case MACH_LEMOTE_ML2F7:
+ ml2f_shutdown();
+ break;
+ case MACH_LEMOTE_YL2F89:
+ yl2f89_shutdown();
+ break;
+ default:
+ break;
+ }
+}
and also, I have replaced that f0 by PCI_MSR_CTRL in this one:
[PATCH v2 4/7] [loongson] lemote-2f: add pci support
about this part in arch/mips/pci/ops-loongson2.c:
+#ifdef CONFIG_CS5536
+ /* cs5536_pci_conf_read4/write4() will call
_rdmsr/_wrmsr() to
+ * access the regsters PCI_MSR_ADDR, PCI_MSR_DATA_LO,
+ * PCI_MSR_DATA_HI, which is bigger than PCI_MSR_CTRL,
so, it
+ * will not go this branch, but the others. so, no
calling dead
+ * loop here.
+ */
+ if ((PCI_IDSEL_CS5536 == device) && (reg <
PCI_MSR_CTRL)) {
If it's hard to find them, I will send the increment patches later.
Regards,
Wu Zhangjin
