OMAP clock and PM architecture fixes for 2.6.32-rc6
Subsystems (5)
Paul Walmsley
maintainercover OMAP clock and PM architecture fixes for 2.6.32-rc6
Hi,
this series contains fixes for some clock and PM architecture components for the 2.6.32-rc series.
- Paul
---
Paul Walmsley (1):
OMAP clock/hwmod: fix off-by-one errors
Rajendra Nayak (1):
OMAP3 clock: Fix the DPLL freqsel computations
Tero Kristo (1):
OMAP3: Clock: Fixed dpll3_m2x2 rate calculation
arch/arm/mach-omap2/clock34xx.c | 2 +- arch/arm/mach-omap2/clock34xx.h | 4 ++-- arch/arm/mach-omap2/cm.c | 2 +- arch/arm/mach-omap2/omap_hwmod.c | 12 +++++------- arch/arm/mach-omap2/prcm.c | 4 ++-- 5 files changed, 11 insertions(+), 13 deletions(-)
Paul Walmsley
maintainerPATCH 1/3 OMAP3 clock: Fix the DPLL freqsel computations
From: Rajendra Nayak <[email protected]>
Fix the freqsel value computation. Use n instead of (n+1)
The formula in the TRM uses a zero-based N, hence the (n+1); however at this point in the clock34xx.c code, N is one-based.
Hayati Bayrakdar <[email protected]> and Nishanth Menon <[email protected]> helped track down this bug.
--- arch/arm/mach-omap2/clock34xx.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
arch/arm/mach-omap2/clock34xx.c +1 -1
@@ -473,7 +473,7 @@ static u16 _omap3_dpll_compute_freqsel(struct clk *clk, u8 n)unsigned long fint;u16 f =0 ; - fint = clk->dpll_data->clk_ref->rate / (n +1 ); + fint = clk->dpll_data->clk_ref->rate / n;pr_debug ("clock: fint is %lu\n" , fint);
Tony Lindgren
maintainerPATCH 1/3 OMAP3 clock: Fix the DPLL freqsel computations
This patch has been applied to the linux-omap
by youw fwiendly patch wobot.
Branch in linux-omap: omap-fixes
Initial commit ID (Likely to change): 8c3b128f6b50f9b8a6b015afeead77fd6de366a6
PatchWorks
http://patchwork.kernel.org/patch/60367/
Git (Likely to change, and takes a while to get mirrored)
http://git.kernel.org/?p=linux/kernel/git/tmlind/linux-omap-2.6.git;a=commit;h=8c3b128f6b50f9b8a6b015afeead77fd6de366a6
Paul Walmsley
maintainerPATCH 2/3 OMAP3: Clock: Fixed dpll3_m2x2 rate calculation
In reply to Paul WalmsleyFrom: Tero Kristo <[email protected]>
Current calculation does not take into account any changes to M2 divisor, and thus when we change VDD2 OPP, dpll3_m2x2 rate does not change. Fixed by re-routing dpll3_m2x2 parent to dpll3_m2.
--- arch/arm/mach-omap2/clock34xx.h | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-)
arch/arm/mach-omap2/clock34xx.h +2 -2
@@ -489,9 +489,9 @@ static struct clk core_ck = {static struct clk dpll3_m2x2_ck = { .name ="dpll3_m2x2_ck" , .ops = &clkops_null, - .parent = &dpll3_x2_ck, + .parent = &dpll3_m2_ck, .clkdm_name ="dpll3_clkdm" , - .recalc = &followparent_recalc, + .recalc = &omap3_clkoutx2_recalc, };/* The PWRDN bit is apparently only available on 3430ES2 and above */
Tony Lindgren
maintainerPATCH 2/3 OMAP3: Clock: Fixed dpll3_m2x2 rate calculation
This patch has been applied to the linux-omap
by youw fwiendly patch wobot.
Branch in linux-omap: omap-fixes
Initial commit ID (Likely to change): 2dcbc9ebbe46c92e748e6180133d10cce69acb4d
PatchWorks
http://patchwork.kernel.org/patch/60368/
Git (Likely to change, and takes a while to get mirrored)
http://git.kernel.org/?p=linux/kernel/git/tmlind/linux-omap-2.6.git;a=commit;h=2dcbc9ebbe46c92e748e6180133d10cce69acb4d
Paul Walmsley
maintainerPATCH 3/3 OMAP clock/hwmod: fix off-by-one errors
In reply to Paul WalmsleyFix loop bailout off-by-one bugs reported by Juha Lepp?nen <[email protected]>.
--- arch/arm/mach-omap2/cm.c | 2 +- arch/arm/mach-omap2/omap_hwmod.c | 12 +++++------- arch/arm/mach-omap2/prcm.c | 4 ++-- 3 files changed, 8 insertions(+), 10 deletions(-)
arch/arm/mach-omap2/cm.c +1 -1
@@ -65,6 +65,6 @@ int omap2_cm_wait_module_ready(s16 prcm_mod, u8 idlest_id, u8 idlest_shift) (i++ < MAX_MODULE_READY_TIME))udelay (1 ); -return (i <MAX_MODULE_READY_TIME ) ?0 : -EBUSY ; +return (i <=MAX_MODULE_READY_TIME ) ?0 : -EBUSY ; }
arch/arm/mach-omap2/omap_hwmod.c +5 -7
@@ -759,14 +759,12 @@ static int _reset(struct omap_hwmod *oh)_write_sysconfig (v, oh); c =0 ; -while (c <MAX_MODULE_RESET_WAIT && - !(omap_hwmod_readl (oh, oh->sysconfig->syss_offs) & -SYSS_RESETDONE_MASK )) { +while (!(omap_hwmod_readl (oh, oh->sysconfig->syss_offs) & +SYSS_RESETDONE_MASK ) && + (c++ <MAX_MODULE_RESET_WAIT ))udelay (1 ); - c++; - } -if (c ==MAX_MODULE_RESET_WAIT ) +if (c >MAX_MODULE_RESET_WAIT )WARN (1 ,"omap_hwmod: %s: failed to reset in %d usec\n" , oh->name,MAX_MODULE_RESET_WAIT );else
@@ -777,7 +775,7 @@ static int _reset(struct omap_hwmod *oh)* _wait_target_ready() or _reset() */ -return (c ==MAX_MODULE_RESET_WAIT ) ? -ETIMEDOUT :0 ; +return (c >MAX_MODULE_RESET_WAIT ) ? -ETIMEDOUT :0 ; }/**
arch/arm/mach-omap2/prcm.c +2 -2
@@ -251,14 +251,14 @@ int omap2_cm_wait_idlest(void __iomem *reg, u32 mask, const char *name) (i++ < MAX_MODULE_ENABLE_WAIT))udelay (1 ); -if (i <MAX_MODULE_ENABLE_WAIT ) +if (i <=MAX_MODULE_ENABLE_WAIT )pr_debug ("cm: Module associated with clock %s ready after %d " "loops\n" , name, i);else pr_err ("cm: Module associated with clock %s didn't enable in " "%d tries\n" , name,MAX_MODULE_ENABLE_WAIT ); -return (i <MAX_MODULE_ENABLE_WAIT ) ?1 :0 ; +return (i <=MAX_MODULE_ENABLE_WAIT ) ?1 :0 ; };void __init omap2_set_globals_prcm (struct omap_globals *omap2_globals)
PATCH 3/3 OMAP clock/hwmod: fix off-by-one errors
Quoted 2 lines · Paul Walmsley
Fix loop bailout off-by-one bugs reported by Juha Lepp?nen <[email protected]>.
I'm not sure the new code is any easier to read than the old code. And what with some checks post-loop being <= and others being >, it's a recipe for cut'n'paste errors happening.
Quoted 13 lines
arch/arm/mach-omap2/omap_hwmod.c +2
-3@@ -759,14 +759,12 @@ static int _reset(struct omap_hwmod *oh)_write_sysconfig (v, oh); c =0 ; -while (c <MAX_MODULE_RESET_WAIT && - !(omap_hwmod_readl (oh, oh->sysconfig->syss_offs) & -SYSS_RESETDONE_MASK )) { + while (!(omap_hwmod_readl(oh, oh->sysconfig->syss_offs) & + SYSS_RESETDONE_MASK) &&
arch/arm/mach-omap2/omap_hwmod.c static int _reset(struct omap_hwmod *oh) · line 770
Quoted 10 lines
arch/arm/mach-omap2/omap_hwmod.c +2
-3+ (c++ < MAX_MODULE_RESET_WAIT))udelay (1 ); - c++; - } -if (c ==MAX_MODULE_RESET_WAIT ) +if (c >MAX_MODULE_RESET_WAIT )WARN (1 ,"omap_hwmod: %s: failed to reset in %d usec\n" , oh->name,MAX_MODULE_RESET_WAIT );else
How about:
for (c = 0; c <= MAX_MODULE_RESET_WAIT; c++) {
if (omap_hwmod_readl(oh, oh->sysconfig->syss_offs) &
SYSS_RESETDONE_MASK) {
pr_debug("omap_hwmod: %s: reset in %d usec\n",
oh->name, c);
return 0;
}
}
WARN(1, "omap_hwmod: %s: failed to reset in %d usec\n",
oh->name, c - 1);
return -ETIMEDOUT;?
Even better would be to turn this into a helper much like wait_event(), which would stop silly mistakes happening.
Paul Walmsley
maintainerPATCH 3/3 OMAP clock/hwmod: fix off-by-one errors
Hi Russell,
Quoted 7 lines · Russell King - ARM Linux
Quoted 2 lines · Paul Walmsley
Fix loop bailout off-by-one bugs reported by Juha Leppänen <[email protected]>.
I'm not sure the new code is any easier to read than the old code. And what with some checks post-loop being <= and others being >, it's a recipe for cut'n'paste errors happening.
...
Quoted 15 lines
How about:
for (c = 0; c <= MAX_MODULE_RESET_WAIT; c++) { if (omap_hwmod_readl(oh, oh->sysconfig->syss_offs) & SYSS_RESETDONE_MASK) { pr_debug("omap_hwmod: %s: reset in %d usec\n", oh->name, c); return 0; } } WARN(1, "omap_hwmod: %s: failed to reset in %d usec\n", oh->name, c - 1); return -ETIMEDOUT;
Quoted 4 lines
?
Even better would be to turn this into a helper much like
wait_event(), which would stop silly mistakes happening.
Agreed. How about this?
arch/arm/plat-omap/include/plat/common.h probably isn't the right place for it, since it's not OMAP-specific...
- Paul
From 972b3bbf2c210086b862aedb35587484bcefe998 Mon Sep 17 00:00:00 2001
From: Paul Walmsley <[email protected]>
Date: Mon, 16 Nov 2009 06:13:15 -0700
Subject: [PATCH] OMAP clock/hwmod: fix off-by-one errors
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Fix loop bailout off-by-one bugs reported by Juha Leppänen <[email protected]>.
This second version incorporates comments from Russell King <[email protected]>. A new macro, 'omap_test_timeout', has been created, with cleaner code, and existing code has been converted to use it.
--- arch/arm/mach-omap2/cm.c | 7 ++++--- arch/arm/mach-omap2/omap_hwmod.c | 13 +++++-------- arch/arm/mach-omap2/prcm.c | 5 ++--- arch/arm/plat-omap/include/plat/common.h | 20 ++++++++++++++++++++ 4 files changed, 31 insertions(+), 14 deletions(-)
arch/arm/mach-omap2/cm.c +4 -3
@@ -21,6 +21,8 @@#include <asm/atomic.h> +#include <plat/common.h> +#include "cm.h" #include "cm-regbits-24xx.h" #include "cm-regbits-34xx.h"
@@ -61,9 +63,8 @@ int omap2_cm_wait_module_ready(s16 prcm_mod, u8 idlest_id, u8 idlest_shift) mask =1 << idlest_shift;/* XXX should be OMAP2 CM */ -while (((cm_read_mod_reg (prcm_mod, cm_idlest_reg) & mask) != ena) && - (i++ <MAX_MODULE_READY_TIME )) -udelay (1 ); +omap_test_timeout (((cm_read_mod_reg (prcm_mod, cm_idlest_reg) & mask) == ena), +MAX_MODULE_READY_TIME , i);return (i <MAX_MODULE_READY_TIME ) ?0 : -EBUSY ; }
arch/arm/mach-omap2/omap_hwmod.c +5 -8
@@ -45,6 +45,7 @@#include <linux/mutex.h> #include <linux/bootmem.h> +#include <plat/common.h> #include <plat/cpu.h> #include <plat/clockdomain.h> #include <plat/powerdomain.h>
@@ -736,7 +737,7 @@ static int _wait_target_ready(struct omap_hwmod *oh)static int _reset (struct omap_hwmod *oh) {u32 r, v; -int c; +int c =0 ; if (!oh->sysconfig || !(oh->sysconfig->sysc_flags & SYSC_HAS_SOFTRESET) ||
@@ -758,13 +759,9 @@ static int _reset(struct omap_hwmod *oh)return r;_write_sysconfig (v, oh); - c =0 ; -while (c <MAX_MODULE_RESET_WAIT && - !(omap_hwmod_readl (oh, oh->sysconfig->syss_offs) & -SYSS_RESETDONE_MASK )) { -udelay (1 ); - c++; - } +omap_test_timeout ((omap_hwmod_readl (oh, oh->sysconfig->syss_offs) & +SYSS_RESETDONE_MASK ), +MAX_MODULE_RESET_WAIT , c);if (c ==MAX_MODULE_RESET_WAIT ) WARN(1, "omap_hwmod: %s: failed to reset in %d usec\n",
arch/arm/mach-omap2/prcm.c +2 -3
@@ -247,9 +247,8 @@ int omap2_cm_wait_idlest(void __iomem *reg, u32 mask, const char *name)BUG ();/* Wait for lock */ -while (((__raw_readl (reg) & mask) != ena) && - (i++ <MAX_MODULE_ENABLE_WAIT )) -udelay (1 ); +omap_test_timeout (((__raw_readl (reg) & mask) == ena), +MAX_MODULE_ENABLE_WAIT , i);if (i <MAX_MODULE_ENABLE_WAIT ) pr_debug("cm: Module associated with clock %s ready after %d "
arch/arm/plat-omap/include/plat/common.h +20 -0
@@ -71,4 +71,24 @@ void omap2_set_globals_sdrc(struct omap_globals *);void omap2_set_globals_control (struct omap_globals *);void omap2_set_globals_prcm (struct omap_globals *); +/** +* omap_test_timeout - busy-loop, testing a condition +* @cond: condition to test until it evaluates to true +* @timeout: maximum number of microseconds in the timeout +* @index: loop index (integer) +* +* Loop waiting for @cond to become true or until at least @timeout +* microseconds have passed. To use, define some integer @index in the +* calling code. After running, if @index == @timeout, then the loop has +* timed out. +*/ +#define omap_test_timeout (cond, timeout, index) \ +({ \ + for (index = 0; index < timeout; index++) { \ + if (cond) \ + break; \ + udelay(1); \ + } \ +}) +#endif /* __ARCH_ARM_MACH_OMAP_COMMON_H */
Signature · 1 line
1.6.5.GIT
PATCH 3/3 OMAP clock/hwmod: fix off-by-one errors
Paul
Quoted 1 line
-----Original Message-----
From: [email protected] [mailto:linux-omap-owner at vger.kernel.org] On Behalf Of Paul
Quoted 5 lines
Walmsley
Sent: Tuesday, November 17, 2009 11:39 AM
To: Russell King - ARM Linux
Cc: linux-arm-kernel at lists.infradead.org; Juha Lepp?nen; linux-omap at vger.kernel.org
Subject: Re: [PATCH 3/3] OMAP clock/hwmod: fix off-by-one errors
<snip>
Quoted 11 lines
arch/arm/plat-omap/include/plat/common.h +10
-0+/** +* omap_test_timeout - busy-loop, testing a condition +* @cond: condition to test until it evaluates to true +* @timeout: maximum number of microseconds in the timeout +* @index: loop index (integer) +* +* Loop waiting for @cond to become true or until at least @timeout +* microseconds have passed. To use, define some integer @index in the +* calling code. After running, if @index == @timeout, then the loop has +* timed out.
arch/arm/plat-omap/include/plat/common.h void omap2_set_globals_sdrc(struct omap_globals *); · line 93
Quoted 10 lines
arch/arm/plat-omap/include/plat/common.h +10
-0+*/ +#define omap_test_timeout (cond, timeout, index) \ +({ \ + for (index = 0; index < timeout; index++) { \ + if (cond) \ + break; \ + udelay(1); \ + } \ +}) +
There is a similar kind of function implemented in USB host, that returns -ETIMEDOUT Which makes more sense for failure case.
Maybe such a generic function for all omaps is needed.
Refer:drivers/usb/host/ehci-hcd.c
static int handshake (struct ehci_hcd *ehci, void __iomem *ptr,
u32 mask, u32 done, int usec)
{
u32 result;
do {
result = ehci_readl(ehci, ptr);
if (result == ~(u32)0) /* card removed */
return -ENODEV;
result &= mask;
if (result == done)
return 0;
udelay (1);
usec--;
} while (usec > 0);
return -ETIMEDOUT;
}Quoted 3 lines
#endif /*
__ARCH_ARM_MACH_OMAP_COMMON_H*/
--
1.6.5.GIT
Tony Lindgren
maintainerPATCH 3/3 OMAP clock/hwmod: fix off-by-one errors
Quoted 41 lines · Pandita, Vikram
Paul
Quoted 7 lines
-----Original Message-----
From: linux-omap-owner at vger.kernel.org [mailto:linux-omap-owner at vger.kernel.org] On Behalf Of Paul Walmsley
Sent: Tuesday, November 17, 2009 11:39 AM
To: Russell King - ARM Linux
Cc: linux-arm-kernel at lists.infradead.org; Juha Lepp?nen; linux-omap at vger.kernel.org
Subject: Re: [PATCH 3/3] OMAP clock/hwmod: fix off-by-one errors<snip>
Quoted 11 lines
arch/arm/plat-omap/include/plat/common.h +10
-0+/** +* omap_test_timeout - busy-loop, testing a condition +* @cond: condition to test until it evaluates to true +* @timeout: maximum number of microseconds in the timeout +* @index: loop index (integer) +* +* Loop waiting for @cond to become true or until at least @timeout +* microseconds have passed. To use, define some integer @index in the +* calling code. After running, if @index == @timeout, then the loop has +* timed out. arch/arm/plat-omap/include/plat/common.h void omap2_set_globals_sdrc(struct omap_globals *); · line 93
Quoted 10 lines
arch/arm/plat-omap/include/plat/common.h +10
-0+*/ +#define omap_test_timeout (cond, timeout, index) \ +({ \ + for (index = 0; index < timeout; index++) { \ + if (cond) \ + break; \ + udelay(1); \ + } \ +}) +There is a similar kind of function implemented in USB host, that returns
-ETIMEDOUTWhich makes more sense for failure case.Maybe such a generic function for all omaps is needed.
Refer:
drivers/usb/host/ehci-hcd.c
static int handshake (structehci_hcd*ehci, void__iomem*ptr,
u32 mask, u32 done, int usec)
Quoted 18 lines
{ u32 result; do { result = ehci_readl(ehci, ptr); if (result == ~(u32)0) /* card removed */ return -ENODEV; result &= mask; if (result == done) return 0; udelay (1); usec--; } while (usec > 0); return -ETIMEDOUT; }Quoted 1 line
#endif /*
__ARCH_ARM_MACH_OMAP_COMMON_H*/
Well we need to stick to minimal changes if we plan to send this in as a fix. Anything more complex we should do the next merge window.
Tony
Tony Lindgren
maintainerPATCH 3/3 OMAP clock/hwmod: fix off-by-one errors
Quoted 59 lines · Tony Lindgren
Quoted 58 lines · Pandita, Vikram
Paul
Quoted 7 lines
-----Original Message-----
From: linux-omap-owner at vger.kernel.org [mailto:linux-omap-owner at vger.kernel.org] On Behalf Of Paul Walmsley
Sent: Tuesday, November 17, 2009 11:39 AM
To: Russell King - ARM Linux
Cc: linux-arm-kernel at lists.infradead.org; Juha Lepp?nen; linux-omap at vger.kernel.org
Subject: Re: [PATCH 3/3] OMAP clock/hwmod: fix off-by-one errors<snip>
Quoted 11 lines
arch/arm/plat-omap/include/plat/common.h +10
-0+/** +* omap_test_timeout - busy-loop, testing a condition +* @cond: condition to test until it evaluates to true +* @timeout: maximum number of microseconds in the timeout +* @index: loop index (integer) +* +* Loop waiting for @cond to become true or until at least @timeout +* microseconds have passed. To use, define some integer @index in the +* calling code. After running, if @index == @timeout, then the loop has +* timed out. arch/arm/plat-omap/include/plat/common.h void omap2_set_globals_sdrc(struct omap_globals *); · line 93
Quoted 10 lines
arch/arm/plat-omap/include/plat/common.h +10
-0+*/ +#define omap_test_timeout (cond, timeout, index) \ +({ \ + for (index = 0; index < timeout; index++) { \ + if (cond) \ + break; \ + udelay(1); \ + } \ +}) +There is a similar kind of function implemented in USB host, that returns
-ETIMEDOUTWhich makes more sense for failure case.Maybe such a generic function for all omaps is needed.
Refer:
drivers/usb/host/ehci-hcd.c
static int handshake (structehci_hcd*ehci, void__iomem*ptr,
u32 mask, u32 done, int usec){ u32 result; do { result = ehci_readl(ehci, ptr); if (result == ~(u32)0) /* card removed */ return -ENODEV; result &= mask; if (result == done) return 0; udelay (1); usec--; } while (usec > 0); return -ETIMEDOUT; }
Quoted 4 lines
Quoted 1 line
Quoted 1 line
#endif /*
__ARCH_ARM_MACH_OMAP_COMMON_H*/Well we need to stick to minimal changes if we plan to send this in as a fix. Anything more complex we should do the next merge window.
To me it sounds like this can wait until the merge window, so not adding to omap-fixes for now.
Regards,
Tony
Paul Walmsley
maintainerPATCH 3/3 OMAP clock/hwmod: fix off-by-one errors
Quoted 2 lines · Tony Lindgren
To me it sounds like this can wait until the merge window, so not adding to omap-fixes for now.
Indeed, there's no rush on this one, so let's set it up for 2.6.33. I'll keep that macro in plat/system.h unless you or Russell have a better suggestion. It might be worth looking at the rest of the Linux codebase to see if there are other potential users; maybe someone might be willing to pick up a "kernel-janitor" task for this?
Vikram, I agree with your comment. Maybe someone can send a patch to clean up ehci-hcd.c and any other users of the same idiom to use this macro for 2.6.33? Looks like also plat-omap/mailbox.c:__mbox_msg_send() and mach-omap2/powerdomain.c:pwrdm_wait_transition() would be good candidates for this, under arch/arm/*omap*.
- Paul