| Message ID | 1361223490-20106-1-git-send-email-hechtb+renesas@gmail.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
Hi Bastian > The struct sys_timer is gone since the patch "ARM: delete struct > sys_timer". Since then we can't use shmobile_timer any longer. Setting > .init_time directly to eva_earlytimer_init fixes the resulting compile > error and makes the code shorter as well. I guess it is 6bb27d7349db51b50c40534710fe164ca0d58902 Can you please add commit number too ? It is easy to check. something like the patch "ARM: delete struct sys_timer" (6bb27d7349db51b50c40534710fe164ca0d58902) > > Signed-off-by: Bastian Hecht <hechtb+renesas@gmail.com> > --- > .../board-armadillo800eva-reference.c | 12 ++---------- > 1 file changed, 2 insertions(+), 10 deletions(-) > > diff --git a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c > index 623754d..ae20979 100644 > --- a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c > +++ b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c > @@ -219,14 +219,6 @@ static void __init eva_earlytimer_init(void) > eva_clock_init(); > } > > -static void __init eva_add_early_devices(void) > -{ > - r8a7740_add_early_devices_dt(); > - > - /* override timer setup with board-specific code */ > - shmobile_timer.init = eva_earlytimer_init; > -} > - > #define RESCNT2 IOMEM(0xe6188020) > static void eva_restart(char mode, const char *cmd) > { > @@ -241,13 +233,13 @@ static const char *eva_boards_compat_dt[] __initdata = { > > DT_MACHINE_START(ARMADILLO800EVA_DT, "armadillo800eva-reference") > .map_io = r8a7740_map_io, > - .init_early = eva_add_early_devices, > + .init_early = r8a7740_add_early_devices_dt, > .init_irq = r8a7740_init_irq_of, > .nr_irqs = NR_IRQS_LEGACY, > .handle_irq = shmobile_handle_irq_intc, > .init_machine = eva_init, > .init_late = shmobile_init_late, > - .init_time = shmobile_timer_init, > + .init_time = eva_earlytimer_init, > .dt_compat = eva_boards_compat_dt, > .restart = eva_restart, > MACHINE_END > -- > 1.7.9.5 > > -- > To unsubscribe from this list: send the line "unsubscribe linux-sh" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html Best regards --- Kuninori Morimoto -- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Feb 18, 2013 at 03:38:10PM -0600, Bastian Hecht wrote: > The struct sys_timer is gone since the patch "ARM: delete struct > sys_timer". Since then we can't use shmobile_timer any longer. Setting > .init_time directly to eva_earlytimer_init fixes the resulting compile > error and makes the code shorter as well. Hi Bastian, Magnus seems to be very keen on the idea that -reference code shouldn't use early timer or early devices if possible. Is it possible to re-work this patch so that armadillo800eva-reference doesn't use early timer? > > Signed-off-by: Bastian Hecht <hechtb+renesas@gmail.com> > --- > .../board-armadillo800eva-reference.c | 12 ++---------- > 1 file changed, 2 insertions(+), 10 deletions(-) > > diff --git a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c > index 623754d..ae20979 100644 > --- a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c > +++ b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c > @@ -219,14 +219,6 @@ static void __init eva_earlytimer_init(void) > eva_clock_init(); > } > > -static void __init eva_add_early_devices(void) > -{ > - r8a7740_add_early_devices_dt(); > - > - /* override timer setup with board-specific code */ > - shmobile_timer.init = eva_earlytimer_init; > -} > - > #define RESCNT2 IOMEM(0xe6188020) > static void eva_restart(char mode, const char *cmd) > { > @@ -241,13 +233,13 @@ static const char *eva_boards_compat_dt[] __initdata = { > > DT_MACHINE_START(ARMADILLO800EVA_DT, "armadillo800eva-reference") > .map_io = r8a7740_map_io, > - .init_early = eva_add_early_devices, > + .init_early = r8a7740_add_early_devices_dt, > .init_irq = r8a7740_init_irq_of, > .nr_irqs = NR_IRQS_LEGACY, > .handle_irq = shmobile_handle_irq_intc, > .init_machine = eva_init, > .init_late = shmobile_init_late, > - .init_time = shmobile_timer_init, > + .init_time = eva_earlytimer_init, > .dt_compat = eva_boards_compat_dt, > .restart = eva_restart, > MACHINE_END > -- > 1.7.9.5 > -- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi Simon, 2013/2/18 Simon Horman <horms@verge.net.au>: > On Mon, Feb 18, 2013 at 03:38:10PM -0600, Bastian Hecht wrote: >> The struct sys_timer is gone since the patch "ARM: delete struct >> sys_timer". Since then we can't use shmobile_timer any longer. Setting >> .init_time directly to eva_earlytimer_init fixes the resulting compile >> error and makes the code shorter as well. > > Hi Bastian, > > Magnus seems to be very keen on the idea that -reference code > shouldn't use early timer or early devices if possible. > > Is it possible to re-work this patch so that armadillo800eva-reference > doesn't use early timer? Hmm I wonder if this is just a naming convention here. If I rename eva_earlytimer_init() to eva_timer_init() the only thing that might disturb us is that we call shmobile_EARLYtimer_init() inside it. But this function only sets a callback named LATE_time_init. So we don't break any timings. So are you ok if I rename eva_earlytimer_init() to eva_timer_init()? Thanks, Bastian >> >> Signed-off-by: Bastian Hecht <hechtb+renesas@gmail.com> >> --- >> .../board-armadillo800eva-reference.c | 12 ++---------- >> 1 file changed, 2 insertions(+), 10 deletions(-) >> >> diff --git a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c >> index 623754d..ae20979 100644 >> --- a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c >> +++ b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c >> @@ -219,14 +219,6 @@ static void __init eva_earlytimer_init(void) >> eva_clock_init(); >> } >> >> -static void __init eva_add_early_devices(void) >> -{ >> - r8a7740_add_early_devices_dt(); >> - >> - /* override timer setup with board-specific code */ >> - shmobile_timer.init = eva_earlytimer_init; >> -} >> - >> #define RESCNT2 IOMEM(0xe6188020) >> static void eva_restart(char mode, const char *cmd) >> { >> @@ -241,13 +233,13 @@ static const char *eva_boards_compat_dt[] __initdata = { >> >> DT_MACHINE_START(ARMADILLO800EVA_DT, "armadillo800eva-reference") >> .map_io = r8a7740_map_io, >> - .init_early = eva_add_early_devices, >> + .init_early = r8a7740_add_early_devices_dt, >> .init_irq = r8a7740_init_irq_of, >> .nr_irqs = NR_IRQS_LEGACY, >> .handle_irq = shmobile_handle_irq_intc, >> .init_machine = eva_init, >> .init_late = shmobile_init_late, >> - .init_time = shmobile_timer_init, >> + .init_time = eva_earlytimer_init, >> .dt_compat = eva_boards_compat_dt, >> .restart = eva_restart, >> MACHINE_END >> -- >> 1.7.9.5 >> -- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Feb 19, 2013 at 09:05:16AM -0600, Bastian Hecht wrote: > Hi Simon, > > 2013/2/18 Simon Horman <horms@verge.net.au>: > > On Mon, Feb 18, 2013 at 03:38:10PM -0600, Bastian Hecht wrote: > >> The struct sys_timer is gone since the patch "ARM: delete struct > >> sys_timer". Since then we can't use shmobile_timer any longer. Setting > >> .init_time directly to eva_earlytimer_init fixes the resulting compile > >> error and makes the code shorter as well. > > > > Hi Bastian, > > > > Magnus seems to be very keen on the idea that -reference code > > shouldn't use early timer or early devices if possible. > > > > Is it possible to re-work this patch so that armadillo800eva-reference > > doesn't use early timer? > > Hmm I wonder if this is just a naming convention here. If I rename > eva_earlytimer_init() to eva_timer_init() the only thing that might > disturb us is that we call shmobile_EARLYtimer_init() inside it. But > this function only sets a callback named LATE_time_init. So we don't > break any timings. > > So are you ok if I rename eva_earlytimer_init() to eva_timer_init()? Perhaps I am missing something. In eva_earlytimer_init() I see a call to shmobile_earlytimer_init() amongst other things. And this seems to be the kind of early timer initialisation that Magnus has asked -reference code not to use. If you look at the setup code in setup-emev2.c and in particular the use and contents emev2_init_delay() then you can see the kind of setup that Magnus is advocating. This is also reflected in "[PATCH v2 0/8] ARM: shmobile: kzm9g reference SMP and cleanup". In particular: * [PATCH v2 5/8] ARM: shmobile: sh73a0: Do not use early devices with DT reference * [PATCH v2 6/8] ARM: shmobile: kzm9g: Do not use early devices -- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Hi Simon, 2013/2/22 Simon Horman <horms@verge.net.au>: > On Tue, Feb 19, 2013 at 09:05:16AM -0600, Bastian Hecht wrote: >> Hi Simon, >> >> 2013/2/18 Simon Horman <horms@verge.net.au>: >> > On Mon, Feb 18, 2013 at 03:38:10PM -0600, Bastian Hecht wrote: >> >> The struct sys_timer is gone since the patch "ARM: delete struct >> >> sys_timer". Since then we can't use shmobile_timer any longer. Setting >> >> .init_time directly to eva_earlytimer_init fixes the resulting compile >> >> error and makes the code shorter as well. >> > >> > Hi Bastian, >> > >> > Magnus seems to be very keen on the idea that -reference code >> > shouldn't use early timer or early devices if possible. >> > >> > Is it possible to re-work this patch so that armadillo800eva-reference >> > doesn't use early timer? >> >> Hmm I wonder if this is just a naming convention here. If I rename >> eva_earlytimer_init() to eva_timer_init() the only thing that might >> disturb us is that we call shmobile_EARLYtimer_init() inside it. But >> this function only sets a callback named LATE_time_init. So we don't >> break any timings. >> >> So are you ok if I rename eva_earlytimer_init() to eva_timer_init()? > > Perhaps I am missing something. > > In eva_earlytimer_init() I see a call to shmobile_earlytimer_init() > amongst other things. And this seems to be the kind of early timer > initialisation that Magnus has asked -reference code not to use. > > If you look at the setup code in setup-emev2.c and in particular > the use and contents emev2_init_delay() then you can see the kind of setup > that Magnus is advocating. > > This is also reflected in "[PATCH v2 0/8] ARM: shmobile: kzm9g reference > SMP and cleanup". In particular: > > * [PATCH v2 5/8] ARM: shmobile: sh73a0: Do not use early devices with DT reference > * [PATCH v2 6/8] ARM: shmobile: kzm9g: Do not use early devices Thanks for guidance to these patches! On which branch does this patchset apply? Or are they even already applied on some publicly available branch? Cheers, Bastian -- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Feb 27, 2013 at 10:32:17AM -0600, Bastian Hecht wrote: > Hi Simon, > > 2013/2/22 Simon Horman <horms@verge.net.au>: > > On Tue, Feb 19, 2013 at 09:05:16AM -0600, Bastian Hecht wrote: > >> Hi Simon, > >> > >> 2013/2/18 Simon Horman <horms@verge.net.au>: > >> > On Mon, Feb 18, 2013 at 03:38:10PM -0600, Bastian Hecht wrote: > >> >> The struct sys_timer is gone since the patch "ARM: delete struct > >> >> sys_timer". Since then we can't use shmobile_timer any longer. Setting > >> >> .init_time directly to eva_earlytimer_init fixes the resulting compile > >> >> error and makes the code shorter as well. > >> > > >> > Hi Bastian, > >> > > >> > Magnus seems to be very keen on the idea that -reference code > >> > shouldn't use early timer or early devices if possible. > >> > > >> > Is it possible to re-work this patch so that armadillo800eva-reference > >> > doesn't use early timer? > >> > >> Hmm I wonder if this is just a naming convention here. If I rename > >> eva_earlytimer_init() to eva_timer_init() the only thing that might > >> disturb us is that we call shmobile_EARLYtimer_init() inside it. But > >> this function only sets a callback named LATE_time_init. So we don't > >> break any timings. > >> > >> So are you ok if I rename eva_earlytimer_init() to eva_timer_init()? > > > > Perhaps I am missing something. > > > > In eva_earlytimer_init() I see a call to shmobile_earlytimer_init() > > amongst other things. And this seems to be the kind of early timer > > initialisation that Magnus has asked -reference code not to use. > > > > If you look at the setup code in setup-emev2.c and in particular > > the use and contents emev2_init_delay() then you can see the kind of setup > > that Magnus is advocating. > > > > This is also reflected in "[PATCH v2 0/8] ARM: shmobile: kzm9g reference > > SMP and cleanup". In particular: > > > > * [PATCH v2 5/8] ARM: shmobile: sh73a0: Do not use early devices with DT reference > > * [PATCH v2 6/8] ARM: shmobile: kzm9g: Do not use early devices > > Thanks for guidance to these patches! On which branch does this > patchset apply? Or are they even already applied on some publicly > available branch? These changes should be present in both next and topic/all+next. -- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
2013/2/28 Simon Horman <horms@verge.net.au>: > On Wed, Feb 27, 2013 at 10:32:17AM -0600, Bastian Hecht wrote: >> Hi Simon, >> >> 2013/2/22 Simon Horman <horms@verge.net.au>: >> > On Tue, Feb 19, 2013 at 09:05:16AM -0600, Bastian Hecht wrote: >> >> Hi Simon, >> >> >> >> 2013/2/18 Simon Horman <horms@verge.net.au>: >> >> > On Mon, Feb 18, 2013 at 03:38:10PM -0600, Bastian Hecht wrote: >> >> >> The struct sys_timer is gone since the patch "ARM: delete struct >> >> >> sys_timer". Since then we can't use shmobile_timer any longer. Setting >> >> >> .init_time directly to eva_earlytimer_init fixes the resulting compile >> >> >> error and makes the code shorter as well. >> >> > >> >> > Hi Bastian, >> >> > >> >> > Magnus seems to be very keen on the idea that -reference code >> >> > shouldn't use early timer or early devices if possible. >> >> > >> >> > Is it possible to re-work this patch so that armadillo800eva-reference >> >> > doesn't use early timer? >> >> >> >> Hmm I wonder if this is just a naming convention here. If I rename >> >> eva_earlytimer_init() to eva_timer_init() the only thing that might >> >> disturb us is that we call shmobile_EARLYtimer_init() inside it. But >> >> this function only sets a callback named LATE_time_init. So we don't >> >> break any timings. >> >> >> >> So are you ok if I rename eva_earlytimer_init() to eva_timer_init()? >> > >> > Perhaps I am missing something. >> > >> > In eva_earlytimer_init() I see a call to shmobile_earlytimer_init() >> > amongst other things. And this seems to be the kind of early timer >> > initialisation that Magnus has asked -reference code not to use. >> > >> > If you look at the setup code in setup-emev2.c and in particular >> > the use and contents emev2_init_delay() then you can see the kind of setup >> > that Magnus is advocating. >> > >> > This is also reflected in "[PATCH v2 0/8] ARM: shmobile: kzm9g reference >> > SMP and cleanup". In particular: >> > >> > * [PATCH v2 5/8] ARM: shmobile: sh73a0: Do not use early devices with DT reference >> > * [PATCH v2 6/8] ARM: shmobile: kzm9g: Do not use early devices >> >> Thanks for guidance to these patches! On which branch does this >> patchset apply? Or are they even already applied on some publicly >> available branch? > > These changes should be present in both next and topic/all+next. Yeah git remote update is my friend... thanks! Ok I've prepared the patches. I think it would make sense to split it into 2 parts and keep the first patch as it is and have a separate one for removing all the early device stuff. I've prepared it in an identical way as you cleaned up the code. If you prefer 1 patch just tell me. Same if you want me to rebase these patches on some specific branch. Cheers, Bastian -- To unsubscribe from this list: send the line "unsubscribe linux-sh" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c index 623754d..ae20979 100644 --- a/arch/arm/mach-shmobile/board-armadillo800eva-reference.c +++ b/arch/arm/mach-shmobile/board-armadillo800eva-reference.c @@ -219,14 +219,6 @@ static void __init eva_earlytimer_init(void) eva_clock_init(); } -static void __init eva_add_early_devices(void) -{ - r8a7740_add_early_devices_dt(); - - /* override timer setup with board-specific code */ - shmobile_timer.init = eva_earlytimer_init; -} - #define RESCNT2 IOMEM(0xe6188020) static void eva_restart(char mode, const char *cmd) { @@ -241,13 +233,13 @@ static const char *eva_boards_compat_dt[] __initdata = { DT_MACHINE_START(ARMADILLO800EVA_DT, "armadillo800eva-reference") .map_io = r8a7740_map_io, - .init_early = eva_add_early_devices, + .init_early = r8a7740_add_early_devices_dt, .init_irq = r8a7740_init_irq_of, .nr_irqs = NR_IRQS_LEGACY, .handle_irq = shmobile_handle_irq_intc, .init_machine = eva_init, .init_late = shmobile_init_late, - .init_time = shmobile_timer_init, + .init_time = eva_earlytimer_init, .dt_compat = eva_boards_compat_dt, .restart = eva_restart, MACHINE_END
The struct sys_timer is gone since the patch "ARM: delete struct sys_timer". Since then we can't use shmobile_timer any longer. Setting .init_time directly to eva_earlytimer_init fixes the resulting compile error and makes the code shorter as well. Signed-off-by: Bastian Hecht <hechtb+renesas@gmail.com> --- .../board-armadillo800eva-reference.c | 12 ++---------- 1 file changed, 2 insertions(+), 10 deletions(-)