[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20111109213157.082651130@clark.kroah.org>
Date: Wed, 09 Nov 2011 13:33:15 -0800
From: Greg KH <gregkh@...e.de>
To: linux-kernel@...r.kernel.org, stable@...r.kernel.org
Cc: torvalds@...ux-foundation.org, akpm@...ux-foundation.org,
alan@...rguk.ukuu.org.uk,
Srinidhi Kasagar <srinidhi.kasagar@...ricsson.com>,
Rabin Vincent <rabin.vincent@...ricsson.com>,
Adrian Bunk <adrian.bunk@...ial.com>,
Linus Walleij <linus.walleij@...aro.org>
Subject: [148/264] ARM: mach-ux500: unlock I&D l2x0 caches before init
3.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Linus Walleij <linus.walleij@...aro.org>
commit 1bf6d2c1bb23533af6930581cc39b74685bc29de upstream.
Apparently U8500 U-Boot versions may leave the l2x0 locked down
before executing the kernel. Make sure we unlock it before we
initialize the l2x0. This fixes a performance problem reported
by Jan Rinze.
The l2x0 core has been modified to unlock the l2x0 by default,
but it will not touch the locking registers if the l2x0 was
already enabled, as on the ux500, so we need this quirk to
make sure it is properly turned off.
Cc: Srinidhi Kasagar <srinidhi.kasagar@...ricsson.com>
Cc: Rabin Vincent <rabin.vincent@...ricsson.com>
Cc: Adrian Bunk <adrian.bunk@...ial.com>
Reported-by: Jan Rinze <janrinze@...il.com>
Tested-by: Robert Marklund <robert.marklund@...ricsson.com>
Signed-off-by: Linus Walleij <linus.walleij@...aro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@...e.de>
---
arch/arm/mach-ux500/cpu.c | 25 ++++++++++++++++++++++++-
1 file changed, 24 insertions(+), 1 deletion(-)
--- a/arch/arm/mach-ux500/cpu.c
+++ b/arch/arm/mach-ux500/cpu.c
@@ -99,7 +99,27 @@ static void ux500_l2x0_inv_all(void)
ux500_cache_sync();
}
-static int ux500_l2x0_init(void)
+static int __init ux500_l2x0_unlock(void)
+{
+ int i;
+
+ /*
+ * Unlock Data and Instruction Lock if locked. Ux500 U-Boot versions
+ * apparently locks both caches before jumping to the kernel. The
+ * l2x0 core will not touch the unlock registers if the l2x0 is
+ * already enabled, so we do it right here instead. The PL310 has
+ * 8 sets of registers, one per possible CPU.
+ */
+ for (i = 0; i < 8; i++) {
+ writel_relaxed(0x0, l2x0_base + L2X0_LOCKDOWN_WAY_D_BASE +
+ i * L2X0_LOCKDOWN_STRIDE);
+ writel_relaxed(0x0, l2x0_base + L2X0_LOCKDOWN_WAY_I_BASE +
+ i * L2X0_LOCKDOWN_STRIDE);
+ }
+ return 0;
+}
+
+static int __init ux500_l2x0_init(void)
{
if (cpu_is_u5500())
l2x0_base = __io_address(U5500_L2CC_BASE);
@@ -108,6 +128,9 @@ static int ux500_l2x0_init(void)
else
ux500_unknown_soc();
+ /* Unlock before init */
+ ux500_l2x0_unlock();
+
/* 64KB way size, 8 way associativity, force WA */
l2x0_init(l2x0_base, 0x3e060000, 0xc0000fff);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists