uboot-mvebu: backport two patches for Marvell A38x

This solves issue with DDR training on Turris Omnia.

Log:
********   DRAM initialization Failed (res 0x1)   ********
DDR3 Training Sequence - FAILED
ERROR ### Please RESET the board ###

Signed-off-by: Josef Schlehofer <pepe.schlehofer@gmail.com>
This commit is contained in:
Josef Schlehofer 2022-01-28 13:39:45 +01:00 committed by Hauke Mehrtens
parent 5c804bc199
commit d16bd89c71
2 changed files with 214 additions and 0 deletions

View File

@ -0,0 +1,116 @@
From 3fc92a215b69ad448c151489228eb340df9a8703 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Marek=20Beh=C3=BAn?= <marek.behun@nic.cz>
Date: Wed, 12 Jan 2022 17:06:59 +0100
Subject: [PATCH] ddr: marvell: a38x: fix SPLIT_OUT_MIX state decision
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
This is a cleaned up and fixed version of a patch
mv_ddr: a380: fix SPLIT_OUT_MIX state decision
in each pattern cycle the bus state can be changed
in order to avoide it, need to back to the same bus state on each
pattern cycle
by
Moti Boskula <motib@marvell.com>
The original patch is not in Marvell's mv-ddr-marvell repository. It was
gives to us by Marvell to fix an issues with DDR training on some
boards, but it cannot be applied as is to mv-ddr-marvell, because it is
a very dirty draft patch that would certainly break other things, mainly
DDR4 training code in mv-ddr-marvell, since it changes common functions.
I have cleaned up the patch and removed stuff that seemed unnecessary
(when removed, it still fixed things). Note that I don't understand
completely what the code does exactly, since I haven't studied the DDR
training code extensively (and I suspect that no one besides some few
people in Marvell understand the code completely).
Anyway after the cleanup the patch still fixes isssues with DDR training
on the failing boards.
There was also a problem with the original patch on some of the Allied
Telesis' x530 boards, reported by Chris Packham. I have asked Chris to
send me some logs, and managed to fix it:
- if you look at the change, you'll notice that it introduces
subtraction of cur_start_win[] and cur_end_win[] members, depending on
a bit set in the current_byte_status variable
- the original patch subtracted cur_start_win[] if either
BYTE_SPLIT_OUT_MIX or BYTE_HOMOGENEOUS_SPLIT_OUT bits were set, but
subtracted cur_end_win[] only if the first one (BYTE_SPLIT_OUT_MIX)
was set
- from Chris Packham logs I discovered that the x530 board where the
original patch introduced DDR training failure, only the
BYTE_HOMOGENEOUS_SPLIT_OUT bit was set, and on our boards where the
patch is needed only the BYTE_SPLIT_OUT_MIX is set in the
current_byte_status variable
- this led me to the hypothesis that both cur_start_win[] and
cur_end_win[] should be subtracted only if BYTE_SPLIT_OUT_MIX bit is
set, the BYTE_HOMOGENEOUS_SPLIT_OUT bit shouldn't be considered at all
- this hypothesis also gains credibility when considering the commit
title ("fix SPLIT_OUT_MIX state decision")
Hopefully this will fix things without breaking anything else.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Stefan Roese <sr@denx.de>
Tested-by: Chris Packham <judge.packham@gmail.com>
---
.../a38x/ddr3_training_centralization.c | 26 +++++++++++++++++++
1 file changed, 26 insertions(+)
--- a/drivers/ddr/marvell/a38x/ddr3_training_centralization.c
+++ b/drivers/ddr/marvell/a38x/ddr3_training_centralization.c
@@ -55,6 +55,7 @@ static int ddr3_tip_centralization(u32 d
enum hws_training_ip_stat training_result[MAX_INTERFACE_NUM];
u32 if_id, pattern_id, bit_id;
u8 bus_id;
+ u8 current_byte_status;
u8 cur_start_win[BUS_WIDTH_IN_BITS];
u8 centralization_result[MAX_INTERFACE_NUM][BUS_WIDTH_IN_BITS];
u8 cur_end_win[BUS_WIDTH_IN_BITS];
@@ -166,6 +167,10 @@ static int ddr3_tip_centralization(u32 d
result[search_dir_id][7]));
}
+ current_byte_status =
+ mv_ddr_tip_sub_phy_byte_status_get(if_id,
+ bus_id);
+
for (bit_id = 0; bit_id < BUS_WIDTH_IN_BITS;
bit_id++) {
/* check if this code is valid for 2 edge, probably not :( */
@@ -174,11 +179,32 @@ static int ddr3_tip_centralization(u32 d
[HWS_LOW2HIGH]
[bit_id],
EDGE_1);
+ if (current_byte_status &
+ BYTE_SPLIT_OUT_MIX) {
+ if (cur_start_win[bit_id] >= 64)
+ cur_start_win[bit_id] -= 64;
+ else
+ cur_start_win[bit_id] = 0;
+ DEBUG_CENTRALIZATION_ENGINE
+ (DEBUG_LEVEL_INFO,
+ ("pattern %d IF %d pup %d bit %d subtract 64 adll from start\n",
+ pattern_id, if_id, bus_id, bit_id));
+ }
cur_end_win[bit_id] =
GET_TAP_RESULT(result
[HWS_HIGH2LOW]
[bit_id],
EDGE_1);
+ if (cur_end_win[bit_id] >= 64 &&
+ (current_byte_status &
+ BYTE_SPLIT_OUT_MIX)) {
+ cur_end_win[bit_id] -= 64;
+ DEBUG_CENTRALIZATION_ENGINE
+ (DEBUG_LEVEL_INFO,
+ ("pattern %d IF %d pup %d bit %d subtract 64 adll from end\n",
+ pattern_id, if_id, bus_id, bit_id));
+ }
+
/* window length */
current_window[bit_id] =
cur_end_win[bit_id] -

View File

@ -0,0 +1,98 @@
From eadc4f512fb43bba2fa4e842c982da919da664be Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Marek=20Beh=C3=BAn?= <marek.behun@nic.cz>
Date: Tue, 4 Jan 2022 15:57:49 +0100
Subject: [PATCH] ddr: marvell: a38x: Fix Synchronous vs Asynchronous mode
determination
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Before commit 4c289425752f ("mv_ddr: a38x: add support for ddr async
mode"), Asynchornous Mode was only used when the CPU Subsystem Clock
Options[4:0] field in the SAR1 register was set to value 0x13: CPU at
2 GHz and DDR at 933 MHz.
Then commit 4c289425752f ("mv_ddr: a38x: add support for ddr async
mode") added support for Asynchornous Modes with frequencies other than
933 MHz (but at least 467 MHz), but the code it added to check for
whether Asynchornous Mode should be used is wrong: it checks whether the
frequency setting in board DDR topology map is set to value other than
MV_DDR_FREQ_SAR.
Thus boards which define a specific value, greater than 400 MHz, for DDR
frequency in their board topology (e.g. Turris Omnia defines
MV_DDR_FREQ_800), are incorrectly put into Asynchornous Mode after that
commit.
The A38x Functional Specification, section 10.12 DRAM Clocking, says:
In Synchornous mode, the DRAM and CPU clocks are edge aligned and run
in 1:2 or 1:3 CPU to DRAM frequency ratios.
Change the check for whether Asynchornous Mode should be used according
to this explanation in Functional Specification.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Tested-by: Chris Packham <judge.packham@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
---
drivers/ddr/marvell/a38x/mv_ddr_plat.c | 19 ++++++++-----------
1 file changed, 8 insertions(+), 11 deletions(-)
--- a/drivers/ddr/marvell/a38x/mv_ddr_plat.c
+++ b/drivers/ddr/marvell/a38x/mv_ddr_plat.c
@@ -167,8 +167,6 @@ static u16 a38x_vco_freq_per_sar_ref_clk
};
-static u32 async_mode_at_tf;
-
static u32 dq_bit_map_2_phy_pin[] = {
1, 0, 2, 6, 9, 8, 3, 7, /* 0 */
8, 9, 1, 7, 2, 6, 3, 0, /* 1 */
@@ -734,7 +732,8 @@ static int ddr3_tip_a38x_set_divider(u8
u32 divider = 0;
u32 sar_val, ref_clk_satr;
u32 async_val;
- u32 freq = mv_ddr_freq_get(frequency);
+ u32 cpu_freq;
+ u32 ddr_freq = mv_ddr_freq_get(frequency);
if (if_id != 0) {
DEBUG_TRAINING_ACCESS(DEBUG_LEVEL_ERROR,
@@ -751,11 +750,14 @@ static int ddr3_tip_a38x_set_divider(u8
ref_clk_satr = reg_read(DEVICE_SAMPLE_AT_RESET2_REG);
if (((ref_clk_satr >> DEVICE_SAMPLE_AT_RESET2_REG_REFCLK_OFFSET) & 0x1) ==
DEVICE_SAMPLE_AT_RESET2_REG_REFCLK_25MHZ)
- divider = a38x_vco_freq_per_sar_ref_clk_25_mhz[sar_val] / freq;
+ cpu_freq = a38x_vco_freq_per_sar_ref_clk_25_mhz[sar_val];
else
- divider = a38x_vco_freq_per_sar_ref_clk_40_mhz[sar_val] / freq;
+ cpu_freq = a38x_vco_freq_per_sar_ref_clk_40_mhz[sar_val];
+
+ divider = cpu_freq / ddr_freq;
- if ((async_mode_at_tf == 1) && (freq > 400)) {
+ if (((cpu_freq % ddr_freq != 0) || (divider != 2 && divider != 3)) &&
+ (ddr_freq > 400)) {
/* Set async mode */
dunit_write(0x20220, 0x1000, 0x1000);
dunit_write(0xe42f4, 0x200, 0x200);
@@ -869,8 +871,6 @@ int ddr3_tip_ext_write(u32 dev_num, u32
int mv_ddr_early_init(void)
{
- struct mv_ddr_topology_map *tm = mv_ddr_topology_map_get();
-
/* FIXME: change this configuration per ddr type
* configure a380 and a390 to work with receiver odt timing
* the odt_config is defined:
@@ -882,9 +882,6 @@ int mv_ddr_early_init(void)
mv_ddr_sw_db_init(0, 0);
- if (tm->interface_params[0].memory_freq != MV_DDR_FREQ_SAR)
- async_mode_at_tf = 1;
-
return MV_OK;
}