Menu
▾
▴
[openocd-svn] Main OpenOCD repository branch, master, updated. v0.4.0-256-g8465e99
|
From: Øyvind H. <go...@us...> - 2010-05-19 07:37:37
|
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "Main OpenOCD repository".
The branch, master has been updated
via 8465e9944291a03a216fa15e0d7ed6eb9d44ba92 (commit)
via b80d0501b66002cba1b3bc97a027d4f79932f20d (commit)
from c86d7bdad4418f4fc3d81a68398187c6480316fa (commit)
Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.
- Log -----------------------------------------------------------------
commit 8465e9944291a03a216fa15e0d7ed6eb9d44ba92
Author: Gary Carlson <gca...@ca...>
Date: Tue May 18 20:59:07 2010 -0700
reset: fix reset halt bug
I was finally able to figure out the cause of this problem. There are two
parts to the patch. The first patch modifies the configuration file I
originally generated for the Atmel AT91SAM9G20 board and achieves the
following:
+++ Splits the reset-init handler into a reset-start handler for some of the
initial configuration activities and keeps the remainder in the reset-init
handler as was the case before. This was the real issue that was causing
the timing problems I identified before. This solution was confirmed with
an o-scope on actual target hardware.
+++ Adds a new instruction in the reset-start handler to disable fast memory
accesses in the reset-start handler. When the target jtag clock is started
out at 2 kHz during system clock initialization, memory writes (i.e.
register write to enable external reset pin -- basically to RSTC_MR) are
naturally slow and cause GDB keep-alive issues (refer to PATCH 2/2 for
additional fixes).
+++ Modifies the configuration file to use srst_only reset action. The
reset-start/reset-init handler split also now allows the correct behavior to
be used in the configuration file (previously had to use both SRST and TRST
even though only SRST is actually used and connected on the evaluation
board).
+++ Adds external NandFlash configuration support to take advantage of flash
driver added earlier. Doesn't fix any bugs but adds functionality that was
marked as TBD before and thrown in when I did other work on the
configuration file.
Signed-off-by: Ãyvind Harboe <oyv...@zy...>
diff --git a/tcl/board/at91sam9g20-ek.cfg b/tcl/board/at91sam9g20-ek.cfg
index c3eb952..b50e8c8 100644
--- a/tcl/board/at91sam9g20-ek.cfg
+++ b/tcl/board/at91sam9g20-ek.cfg
@@ -13,16 +13,15 @@
# the AT91SAM9260 and shares the same tap ID as it.
set _CHIPNAME at91sam9g20
+set _FLASHTYPE nandflash_cs3
set _ENDIAN little
set _CPUTAPID 0x0792603f
-# Set reset type. Note that the AT91SAM9G20-EK board has the trst signal disconnected. In theory this script
-# therefore should require "srst_only". With some J-Link debuggers at least, "srst_only" causes a temporary USB
-# communication fault. This appears to be more likely attributed to an internal proprietary firmware quirk inside the
-# dongle itself. Using "trst_and_srst" works fine, however. So if you can't beat them -- join them. If you are using
-# something other the a J-Link dongle you may be able to change this back to "srst_only".
+# Set reset type. Note that the AT91SAM9G20-EK board has the trst signal disconnected. Therefore
+# the reset needs to be configured for "srst_only". If for some reason, a zero-ohm jumper is
+# added to the board to connect the trst signal, then this parameter may need to be changed.
-reset_config trst_and_srst
+reset_config srst_only
# Set up the CPU and generate a new jtag tap for AT91SAM9G20.
@@ -55,10 +54,16 @@ $_TARGETNAME configure -work-area-phys 0x00300000 -work-area-size 0x4000 -work-a
# an event handler where these special activities can take place.
scan_chain
-$_TARGETNAME configure -event reset-init {at91sam9g20_init}
+$_TARGETNAME configure -event reset-init {at91sam9g20_reset_init}
+$_TARGETNAME configure -event reset-start {at91sam9g20_reset_start}
# NandFlash configuration and definition
-# Future TBD
+
+nand device nandflash_cs3 at91sam9 $_TARGETNAME 0x40000000 0xfffffe800
+at91sam9 cle 0 22
+at91sam9 ale 0 21
+at91sam9 rdy_busy 0 0xfffff800 13
+at91sam9 ce 0 0xfffff800 14
proc read_register {register} {
set result ""
@@ -66,7 +71,22 @@ proc read_register {register} {
return $result(0)
}
-proc at91sam9g20_init { } {
+proc at91sam9g20_reset_start { } {
+
+ # Make sure that the the jtag is running slow, since there are a number of different ways the board
+ # can be configured coming into this state that can cause communication problems with the jtag
+ # adapter. Also since this call can be made following a "reset init" where fast memory accesses
+ # are enabled, need to temporarily shut this down so that the RSTC_MR register can be written at slower
+ # jtag speed without causing GDB keep alive problem.
+
+ arm7_9 fast_memory_access disable
+ adapter_khz 2 # Slow-speed oscillator enabled at reset, so run jtag speed slow.
+ halt # Make sure processor is halted, or error will result in following steps.
+ wait_halt 10000
+ mww 0xfffffd08 0xa5000501 # RSTC_MR : enable user reset.
+}
+
+proc at91sam9g20_reset_init { } {
# At reset AT91SAM9G20 chip runs on slow clock (32.768 kHz). To shift over to a normal clock requires
# a number of steps that must be carefully performed. The process outline below follows the
@@ -77,9 +97,6 @@ proc at91sam9g20_init { } {
# means the master clock (MCLK) must be at or below 133 MHz or timing errors will occur. The processor
# core can operate up to 400 MHz and therefore PCLK must be at or below this to function properly.
- adapter_khz 2 # Slow-speed oscillator enabled at reset, so run jtag speed slow.
- halt # Make sure processor is halted, or error will result in following steps.
- mww 0xfffffd08 0xa5000501 # RSTC_MR : enable user reset.
mww 0xfffffd44 0x00008000 # WDT_MR : disable watchdog.
# Enable the main 18.432 MHz oscillator in CKGR_MOR register.
@@ -114,9 +131,10 @@ proc at91sam9g20_init { } {
adapter_khz 0
- # Enable faster DCC downloads.
+ # Enable faster DCC downloads and memory accesses.
arm7_9 dcc_downloads enable
+ arm7_9 fast_memory_access enable
# To be able to use external SDRAM, several peripheral configuration registers must
# be modified. The first change is made to PIO_ASR to select peripheral functions
@@ -134,16 +152,34 @@ proc at91sam9g20_init { } {
# The AT91SAM9G20-EK evaluation board has built-in NandFlash. The exact physical timing characteristics
# for the memory type used on the current board (MT29F2G08AACWP) can be established by setting
- # four registers in order: SMC_SETUP3, SMC_PULSE3, SMC_CYCLE3, and SMC_MODE3.
-
- mww 0xffffec30 0x00020002
- mww 0xffffec34 0x04040404
- mww 0xffffec38 0x00070007
- mww 0xffffec3c 0x00030003
-
- # Identify NandFlash bank 0. Disabled at the moment because a memory driver is not yet complete.
-
-# nand probe 0
+ # a number of registers. The first step involves setting up the general I/O pins on the processor
+ # to be able to interface and support the external memory.
+
+ mww 0xfffffc10 0x00000010 # PMC_PCER : enable PIOC clock
+ mww 0xfffff800 0x00006000 # PIOC_PER : enable PIO function for 13(RDY/~BSY) and 14(~CS)
+ mww 0xfffff810 0x00004000 # PIOC_OER : enable output on 14
+ mww 0xfffff814 0x00002000 # PIOC_ODR : disable output on 13
+ mww 0xfffff830 0x00004000 # PIOC_SODR : set 14 to disable NAND
+
+ # The exact physical timing characteristics for the memory type used on the current board
+ # (MT29F2G08AACWP) can be established by setting four registers in order: SMC_SETUP3,
+ # SMC_PULSE3, SMC_CYCLE3, and SMC_MODE3. Computing the exact values of these registers
+ # is a little tedious to do here. If you have questions about how to do this, Atmel has
+ # a decent application note #6255B that covers this process.
+
+ mww 0xffffec30 0x00020002 # SMC_SETUP3 : 2 clock cycle setup for NRD and NWE
+ mww 0xffffec34 0x04040404 # SMC_PULSE3 : 4 clock cycle pulse for all signals
+ mww 0xffffec38 0x00070006 # SMC_CYCLE3 : 7 clock cycle NRD and 6 NWE cycle
+ mww 0xffffec3C 0x00020003 # SMC_MODE3 : NRD and NWE control, no NWAIT, 8-bit DBW,
+
+ mww 0xffffe800 0x00000001 # ECC_CR : reset the ECC parity registers
+ mww 0xffffe804 0x00000002 # ECC_MR : page size is 2112 words (word is 8 bits)
+
+ # Identify NandFlash bank 0.
+
+ nand probe nandflash_cs3
+
+ # The AT91SAM9G20-EK evaluation board has build-in serial data flash also.
# Now setup SDRAM. This is tricky and configuration is very important for reliability! The current calculations
# are based on 2 x Micron MT48LC16M16A2-75 memory (4 M x 16 bit x 4 banks). If you use this file as a reference
commit b80d0501b66002cba1b3bc97a027d4f79932f20d
Author: Gary Carlson <gca...@ca...>
Date: Tue May 18 20:59:00 2010 -0700
target: slow targets could cause GDB to time out
This second half of the patch is proposed to clean up some GDB keep alive
issues on arm7_9 targets that start up with very slow clocks. If an attempt
is made to write to key registers on the processor with a slow jtag speed,
GDB timeout warnings appear on the console (at least mine) when "reset halt"
or "reset init" commands are issued from the gdb client:
*** BEFORE PATCH ***
(gdb) monitor reset init
fast memory access is disabled
2 kHz
keep_alive() was not invoked in the 1000ms timelimit. GDB alive packet not
sent! (1026). Workaround: increase "set remotetimeout" in GDB
JTAG tap: at91sam9g20.cpu tap/device found: 0x0792603f (mfg: 0x01f, part:
0x7926, ver: 0x0)
target state: halted
target halted in ARM state due to breakpoint, current mode: Supervisor
cpsr: 0x000000d3 pc: 0x00000000
MMU: disabled, D-Cache: disabled, I-Cache: disabled
keep_alive() was not invoked in the 1000ms timelimit. GDB alive packet not
sent! (1027). Workaround: increase "set remotetimeout" in GDB
keep_alive() was not invoked in the 1000ms timelimit. GDB alive packet not
sent! (1006). Workaround: increase "set remotetimeout" in GDB
keep_alive() was not invoked in the 1000ms timelimit. GDB alive packet not
sent! (1006). Workaround: increase "set remotetimeout" in GDB
keep_alive() was not invoked in the 1000ms timelimit. GDB alive packet not
sent! (1006). Workaround: increase "set remotetimeout" in GDB
keep_alive() was not invoked in the 1000ms timelimit. GDB alive packet not
sent! (1004). Workaround: increase "set remotetimeout" in GDB
RCLK - adaptive
dcc downloads are enabled
fast memory access is enabled
NAND flash device 'NAND 256MiB 3,3V 8-bit' found
(gdb)
I added additional keep alive steps in areas that troubleshooting revealed
were causing problems. I only did this however for non-fast write memory
accesses. I don't think most people would be using fast memory accesses to
write to memory when the jtag and system clocks are slow anyway.
If you disagree with my feeling, think there is a more elegant way to handle
the problem, or think the patch will cause other unforeseen problems with
other targets, let me know. As you can see below, the patch does eliminate
the problem on my development station and I suspect that it will benefit
others.
*** AFTER PATCH ***
(gdb) monitor reset init
fast memory access is disabled
2 kHz
JTAG tap: at91sam9g20.cpu tap/device found: 0x0792603f (mfg: 0x01f, part:
0x7926, ver: 0x0)
target state: halted
target halted in ARM state due to breakpoint, current mode: Supervisor
cpsr: 0x000000d3 pc: 0x00000000
MMU: disabled, D-Cache: disabled, I-Cache: disabled
RCLK - adaptive
dcc downloads are enabled
fast memory access is enabled
NAND flash device 'NAND 256MiB 3,3V 8-bit' found
(gdb)
Gary Carlson
Gary Carlson, MSEE
Principal Engineer
Carlson-Minot Inc.
diff --git a/src/target/arm7_9_common.c b/src/target/arm7_9_common.c
index ac18b37..abe0c2f 100644
--- a/src/target/arm7_9_common.c
+++ b/src/target/arm7_9_common.c
@@ -2463,7 +2463,20 @@ int arm7_9_write_memory(struct target *target, uint32_t address, uint32_t size,
if (arm7_9->fast_memory_access)
retval = arm7_9_execute_fast_sys_speed(target);
else
+ {
retval = arm7_9_execute_sys_speed(target);
+
+ /*
+ * if memory writes are made when the clock is running slow
+ * (i.e. 32 kHz) which is necessary in some scripts to reconfigure
+ * processor operations after a "reset halt" or "reset init",
+ * need to immediately stroke the keep alive or will end up with
+ * gdb "keep alive not sent error message" problem.
+ */
+
+ keep_alive();
+ }
+
if (retval != ERROR_OK)
{
return retval;
@@ -2499,7 +2512,20 @@ int arm7_9_write_memory(struct target *target, uint32_t address, uint32_t size,
if (arm7_9->fast_memory_access)
retval = arm7_9_execute_fast_sys_speed(target);
else
+ {
retval = arm7_9_execute_sys_speed(target);
+
+ /*
+ * if memory writes are made when the clock is running slow
+ * (i.e. 32 kHz) which is necessary in some scripts to reconfigure
+ * processor operations after a "reset halt" or "reset init",
+ * need to immediately stroke the keep alive or will end up with
+ * gdb "keep alive not sent error message" problem.
+ */
+
+ keep_alive();
+ }
+
if (retval != ERROR_OK)
{
return retval;
@@ -2534,7 +2560,20 @@ int arm7_9_write_memory(struct target *target, uint32_t address, uint32_t size,
if (arm7_9->fast_memory_access)
retval = arm7_9_execute_fast_sys_speed(target);
else
- retval = arm7_9_execute_sys_speed(target);
+ {
+ retval = arm7_9_execute_sys_speed(target);
+
+ /*
+ * if memory writes are made when the clock is running slow
+ * (i.e. 32 kHz) which is necessary in some scripts to reconfigure
+ * processor operations after a "reset halt" or "reset init",
+ * need to immediately stroke the keep alive or will end up with
+ * gdb "keep alive not sent error message" problem.
+ */
+
+ keep_alive();
+ }
+
if (retval != ERROR_OK)
{
return retval;
-----------------------------------------------------------------------
Summary of changes:
src/target/arm7_9_common.c | 41 ++++++++++++++++++++-
tcl/board/at91sam9g20-ek.cfg | 82 ++++++++++++++++++++++++++++++------------
2 files changed, 99 insertions(+), 24 deletions(-)
hooks/post-receive
--
Main OpenOCD repository
|
