| Message ID | 1423059315-28519-3-git-send-email-geert+renesas@glider.be (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wednesday 04 February 2015 15:15:12 Geert Uytterhoeven wrote: > DT_MACHINE_START(R8A7740_DT, "Generic R8A7740 (Flattened Device Tree)") > + .l2c_aux_val = L2C_AUX_CTRL_SHARED_OVERRIDE, > + .l2c_aux_mask = ~L2C_AUX_CTRL_SHARED_OVERRIDE, > .map_io = r8a7740_map_io, > .init_early = shmobile_init_delay, > .init_irq = r8a7740_init_irq_of, +Russell I'd hope we could avoid having any overrides in here that are not specified in DT. I can never remember what we discussed about particular bits in the past though. Is this bit something we could add a binding for, or could we make it enabled by default? I assume you have to add it because the boot loader sets the wrong default, right? Arnd
Hi Arnd, On Wed, Feb 4, 2015 at 3:34 PM, Arnd Bergmann <arnd@arndb.de> wrote: > On Wednesday 04 February 2015 15:15:12 Geert Uytterhoeven wrote: >> DT_MACHINE_START(R8A7740_DT, "Generic R8A7740 (Flattened Device Tree)") >> + .l2c_aux_val = L2C_AUX_CTRL_SHARED_OVERRIDE, >> + .l2c_aux_mask = ~L2C_AUX_CTRL_SHARED_OVERRIDE, >> .map_io = r8a7740_map_io, >> .init_early = shmobile_init_delay, >> .init_irq = r8a7740_init_irq_of, > > +Russell > > I'd hope we could avoid having any overrides in here that are not > specified in DT. I can never remember what we discussed about particular > bits in the past though. Is this bit something we could add a binding > for, or could we make it enabled by default? > > I assume you have to add it because the boot loader sets the wrong > default, right? Indeed. Note that I only try to preserve the register value before and after migrating to generic l2c OF initialization. I did read (part of) the pl310 documentation, but it's still not clear to me if we really need L2C_AUX_CTRL_SHARED_OVERRIDE. Perhaps this was just copied from somewhere else without much afterthought? The previous version was in thread http://lists.infradead.org/pipermail/linux-arm-kernel/2015-January/318328.html Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
On Wednesday 04 February 2015 15:43:43 Geert Uytterhoeven wrote: > > Indeed. Note that I only try to preserve the register value before and after > migrating to generic l2c OF initialization. I did read (part of) the pl310 > documentation, but it's still not clear to me if we really need > L2C_AUX_CTRL_SHARED_OVERRIDE. Perhaps this was just copied from > somewhere else without much afterthought? Have you checked what the hardware/bootloader actually sets in that bit? Maybe you can remove that override completely and the initial Linux default was just set to the same that is already set on powerup? Arnd
Hi Arnd, On Wed, Feb 4, 2015 at 5:36 PM, Arnd Bergmann <arnd@arndb.de> wrote: > On Wednesday 04 February 2015 15:43:43 Geert Uytterhoeven wrote: >> >> Indeed. Note that I only try to preserve the register value before and after >> migrating to generic l2c OF initialization. I did read (part of) the pl310 >> documentation, but it's still not clear to me if we really need >> L2C_AUX_CTRL_SHARED_OVERRIDE. Perhaps this was just copied from >> somewhere else without much afterthought? > > Have you checked what the hardware/bootloader actually sets in that bit? Yes, it boots with 0x02040000, i.e. L2C_AUX_CTRL_SHARED_OVERRIDE (bit 22) is not set. After the l2c driver is finished with it, it contains 0x46440001. > Maybe you can remove that override completely and the initial Linux > default was just set to the same that is already set on powerup? Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
On Wednesday 04 February 2015 18:04:57 Geert Uytterhoeven wrote: > On Wed, Feb 4, 2015 at 5:36 PM, Arnd Bergmann <arnd@arndb.de> wrote: > > On Wednesday 04 February 2015 15:43:43 Geert Uytterhoeven wrote: > >> > >> Indeed. Note that I only try to preserve the register value before and after > >> migrating to generic l2c OF initialization. I did read (part of) the pl310 > >> documentation, but it's still not clear to me if we really need > >> L2C_AUX_CTRL_SHARED_OVERRIDE. Perhaps this was just copied from > >> somewhere else without much afterthought? > > > > Have you checked what the hardware/bootloader actually sets in that bit? > > Yes, it boots with 0x02040000, i.e. L2C_AUX_CTRL_SHARED_OVERRIDE > (bit 22) is not set. > > After the l2c driver is finished with it, it contains 0x46440001. I see. I guess someone understand the effect of this flag. Does it make a difference on the machine you are looking at? I've read the spec at http://infocenter.arm.com/help/topic/com.arm.doc.ddi0246a/DDI0246A_l2cc_pl310_r0p0_trm.pdf on the topic but couldn't make sense of it to understand whether a particular setting might be required for specific hardware or might be a performance optimization. Arnd
Hi Arnd, On Wed, Feb 4, 2015 at 7:33 PM, Arnd Bergmann <arnd@arndb.de> wrote: > On Wednesday 04 February 2015 18:04:57 Geert Uytterhoeven wrote: >> On Wed, Feb 4, 2015 at 5:36 PM, Arnd Bergmann <arnd@arndb.de> wrote: >> > On Wednesday 04 February 2015 15:43:43 Geert Uytterhoeven wrote: >> >> >> >> Indeed. Note that I only try to preserve the register value before and after >> >> migrating to generic l2c OF initialization. I did read (part of) the pl310 >> >> documentation, but it's still not clear to me if we really need >> >> L2C_AUX_CTRL_SHARED_OVERRIDE. Perhaps this was just copied from >> >> somewhere else without much afterthought? >> > >> > Have you checked what the hardware/bootloader actually sets in that bit? >> >> Yes, it boots with 0x02040000, i.e. L2C_AUX_CTRL_SHARED_OVERRIDE >> (bit 22) is not set. >> >> After the l2c driver is finished with it, it contains 0x46440001. > > I see. I guess someone understand the effect of this flag. > Does it make a difference on the machine you are looking at? Nope. I booted it without, too. No obvious differences. > I've read the spec at http://infocenter.arm.com/help/topic/com.arm.doc.ddi0246a/DDI0246A_l2cc_pl310_r0p0_trm.pdf > on the topic but couldn't make sense of it to understand whether a > particular setting might be required for specific hardware or might > be a performance optimization. Thanks! Ah, so I'm not alone... Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds
diff --git a/arch/arm/mach-shmobile/setup-r8a7740.c b/arch/arm/mach-shmobile/setup-r8a7740.c index d191cf4197313482..3ebd4dfc8e853a9b 100644 --- a/arch/arm/mach-shmobile/setup-r8a7740.c +++ b/arch/arm/mach-shmobile/setup-r8a7740.c @@ -835,10 +835,6 @@ static void __init r8a7740_generic_init(void) { r8a7740_meram_workaround(); -#ifdef CONFIG_CACHE_L2X0 - /* Shared attribute override enable, 32K*8way */ - l2x0_init(IOMEM(0xf0002000), 0x00400000, 0xc20f0fff); -#endif of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL); } @@ -855,6 +851,8 @@ static const char *r8a7740_boards_compat_dt[] __initdata = { }; DT_MACHINE_START(R8A7740_DT, "Generic R8A7740 (Flattened Device Tree)") + .l2c_aux_val = L2C_AUX_CTRL_SHARED_OVERRIDE, + .l2c_aux_mask = ~L2C_AUX_CTRL_SHARED_OVERRIDE, .map_io = r8a7740_map_io, .init_early = shmobile_init_delay, .init_irq = r8a7740_init_irq_of,
Migrate the generic r8a7740 platform from calling l2x0_of_init() to the generic l2c OF initialization. Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> --- v2 - Keep only {,~}L2C_AUX_CTRL_SHARED_OVERRIDE --- arch/arm/mach-shmobile/setup-r8a7740.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)