[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1362978365-5593-8-git-send-email-lokeshvutla@ti.com>
Date: Mon, 11 Mar 2013 10:36:04 +0530
From: Lokesh Vutla <lokeshvutla@...com>
To: <linux-kernel@...r.kernel.org>, <santosh.shilimkar@...com>,
<gregkh@...uxfoundation.org>
CC: <linux-omap@...r.kernel.org>,
<linux-arm-kernel@...ts.infradead.org>,
Grygorii Strashko <grygorii.strashko@...com>,
Vitaly Chernooky <vitaly.chernooky@...com>,
Oleksandr Dmytryshyn <oleksandr.dmytryshyn@...com>,
Lokesh Vutla <lokeshvutla@...com>
Subject: [PATCH 7/8] memory: emif: errata i743: Prohibit usage of Power-Down mode
From: Grygorii Strashko <grygorii.strashko@...com>
ERRATA DESCRIPTION :
The EMIF supports power-down state for low power. The EMIF
automatically puts the SDRAM into power-down after the memory is
not accessed for a defined number of cycles and the
EMIF_PWR_MGMT_CTRL[10:8] REG_LP_MODE bit field is set to 0x4.
As the EMIF supports automatic output impedance calibration, a ZQ
calibration long command is issued every time it exits active
power-down and precharge power-down modes. The EMIF waits and
blocks any other command during this calibration.
The EMIF does not allow selective disabling of ZQ calibration upon
exit of power-down mode. Due to very short periods of power-down
cycles, ZQ calibration overhead creates bandwidth issues and
increases overall system power consumption. On the other hand,
issuing ZQ calibration long commands when exiting self-refresh is
still required.
WORKAROUND :
Because there is no power consumption benefit of the power-down due
to the calibration and there is a performance risk, the guideline
is to not allow power-down state and, therefore, to not have set
the EMIF_PWR_MGMT_CTRL[10:8] REG_LP_MODE bit field to 0x4.
Signed-off-by: Grygorii Strashko <grygorii.strashko@...com>
Signed-off-by: Vitaly Chernooky <vitaly.chernooky@...com>
Signed-off-by: Oleksandr Dmytryshyn <oleksandr.dmytryshyn@...com>
Signed-off-by: Lokesh Vutla <lokeshvutla@...com>
---
drivers/memory/emif.c | 35 +++++++++++++++++++++++++++++++++++
1 file changed, 35 insertions(+)
diff --git a/drivers/memory/emif.c b/drivers/memory/emif.c
index f75806a..119503a 100644
--- a/drivers/memory/emif.c
+++ b/drivers/memory/emif.c
@@ -257,6 +257,41 @@ static void set_lpmode(struct emif_data *emif, u8 lpmode)
u32 temp;
void __iomem *base = emif->base;
+ /*
+ * Workaround for errata i743 - LPDDR2 Power-Down State is Not
+ * Efficient
+ *
+ * i743 DESCRIPTION:
+ * The EMIF supports power-down state for low power. The EMIF
+ * automatically puts the SDRAM into power-down after the memory is
+ * not accessed for a defined number of cycles and the
+ * EMIF_PWR_MGMT_CTRL[10:8] REG_LP_MODE bit field is set to 0x4.
+ * As the EMIF supports automatic output impedance calibration, a ZQ
+ * calibration long command is issued every time it exits active
+ * power-down and precharge power-down modes. The EMIF waits and
+ * blocks any other command during this calibration.
+ * The EMIF does not allow selective disabling of ZQ calibration upon
+ * exit of power-down mode. Due to very short periods of power-down
+ * cycles, ZQ calibration overhead creates bandwidth issues and
+ * increases overall system power consumption. On the other hand,
+ * issuing ZQ calibration long commands when exiting self-refresh is
+ * still required.
+ *
+ * WORKAROUND
+ * Because there is no power consumption benefit of the power-down due
+ * to the calibration and there is a performance risk, the guideline
+ * is to not allow power-down state and, therefore, to not have set
+ * the EMIF_PWR_MGMT_CTRL[10:8] REG_LP_MODE bit field to 0x4.
+ */
+ if ((emif->plat_data->ip_rev == EMIF_4D) &&
+ (EMIF_LP_MODE_PWR_DN == lpmode)) {
+ WARN_ONCE(1,
+ "REG_LP_MODE = LP_MODE_PWR_DN(4) is prohibited by"
+ "erratum i743 switch to LP_MODE_SELF_REFRESH(2)\n");
+ /* rallback LP_MODE to Self-refresh mode */
+ lpmode = EMIF_LP_MODE_SELF_REFRESH;
+ }
+
temp = readl(base + EMIF_POWER_MANAGEMENT_CONTROL);
temp &= ~LP_MODE_MASK;
temp |= (lpmode << LP_MODE_SHIFT);
--
1.7.9.5
--
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