Darjeeling updates

.gitlab-ci.d/opentitan/qemu-ot.yml

@@ -1,5 +1,5 @@
 variables:
- BAREMETAL_REF: "240606-1"
+ BAREMETAL_REF: "240703-1"
  QEMU_BUILD_OPTS: ""
 
 include:

docs/opentitan/darjeeling.md

@@ -135,7 +135,8 @@ See [`tools.md`](tools.md)
  is set to 10 MHz. This option is very useful/mandatory to run many OpenTitan tests that rely on
  time or CPU cycle to validate features. Using `-icount` option slows down execution speed though,
  so it is not recommended to use it when the main goal is to develop SW to run on the virtual
- machine.
+ machine. An alternative is to use `-icount shift=auto`, which offers fatest emulation execution,
+ while preserving an accurate ratio between the vCPU clock and the virtual devices.
 
 * `no_epmp_cfg=true` can be appended to the machine option switch, _i.e._
  `-M ot-darjeeeling,no_epmp_cfg=true` to disable the initial ePMP configuration, which can be very
@@ -162,6 +163,10 @@ See [`tools.md`](tools.md)
 * `-cpu lowrisc-ibex,x-zbr=false` can be used to force disable the Zbr experimental-and-deprecated
  RISC-V bitmap extension for CRC32 extension.
 
+* `-global ot-rstmgr.fatal_reset=N`, where `N` is an unsigned integer. Force QEMU VM to exit the
+ N^th^ time the reset manager received a reset request, rather than rebooting the whole machine as
+ the default behavior.
+
 ### AES
 
 * `-global ot-aes.fast-mode=false` can be used to better emulate AES HW IP, as some OT tests expect

docs/opentitan/dtm.md

@@ -1,14 +1,14 @@
 # `dtm.py`
 
-`dtm.py` checks that the JTAG/DTM/DM stack is up and running and demonstrates how to use the
-Debug Module to access the Ibex core.
+`dtm.py` checks that the JTAG/DTM/DM stack is up and running and demonstrates how to use the Debug
+Module to access the Ibex core.
 
 ## Usage
 
 ````text
-usage: dtm.py [-h] [-H HOST] [-P PORT] [-Q] [-l IR_LENGTH] [-b BASE] [-I] [-c]
- [-C MISA_CHECK] [-x] [-X] [-a ADDRESS] [-m {read,write}]
- [-s SIZE] [-f FILE] [-e ELF] [-v] [-d]
+usage: dtm.py [-h] [-H HOST] [-P PORT] [-Q] [-t] [-l IR_LENGTH] [-b BASE] [-I]
+ [-c CSR] [-C CSR_CHECK] [-x] [-X] [-a ADDRESS] [-m {read,write}]
+ [-s SIZE] [-f FILE] [-e ELF] [-F] [-v] [-d]
 
 Debug Transport Module tiny demo
 
@@ -18,12 +18,12 @@ options:
 Virtual machine:
  -H HOST, --host HOST JTAG host (default: localhost)
  -P PORT, --port PORT JTAG port, default: 3335
- -Q, --no-quit  do not ask the QEMU to quit on exit
+ -t, --terminate terminate QEMU when done
 
 DMI:
  -l IR_LENGTH, --ir-length IR_LENGTH
- bit length of the IR register
- -b BASE, --base BASE define DMI base address
+ bit length of the IR register (default: 5)
+ -b BASE, --base BASE define DMI base address (default: 0x0)
 
 Info:
  -I, --info report JTAG ID code and DTM configuration
@@ -85,13 +85,13 @@ Extras:
 * `-P` specify the TCP port of the JTAG server in the QEMU VM, should follow the TCP setting of the
  `-chardev socket,id=taprbb,...` option for invoking QEMU.
 
-* `-Q` do not send QEMU a request for termination when this script exits.
-
 * `-s` specify the number of bytes to read from or write to memory. Useful with the `--mem` option.
  See also the `--address` option. This option may be omitted for the `write` memory operation, in
  which case the size of the specified file is used. Note that only sizes multiple of 4-byte are
  supported for now.
 
+* `-t` send QEMU a request for termination when this script exits.
+
 * `-v` can be repeated to increase verbosity of the script, mostly for debug purpose.
 
 * `-X` do not attempt to resume normal execution of the hart once DTM operation have been completed.

docs/opentitan/pyot.md

@@ -5,12 +5,13 @@
 ## Usage
 
 ````text
-usage: pyot.py [-h] [-D DELAY] [-i ICOUNT] [-L LOG_FILE] [-M LOG] [-m MACHINE]
- [-Q OPTS] [-q QEMU] [-p DEVICE] [-t TRACE] [-S FIRST_SOC] [-s]
- [-U] [-b file] [-c JSON] [-e] [-f RAW] [-K] [-l file] [-O RAW]
- [-o VMEM] [-r ELF] [-w CSV] [-x file] [-X] [-F TEST]
- [-k SECONDS] [-z] [-R] [-T FACTOR] [-Z] [-v] [-d] [--log-time]
- [--debug LOGGER] [--info LOGGER] [--warn LOGGER]
+usage: pyot.py [-h] [-D DELAY] [-i ICOUNT] [-L LOG_FILE] [-M VARIANT] [-N LOG]
+ [-m MACHINE] [-Q OPTS] [-q QEMU] [-P VCP] [-p DEVICE]
+ [-t TRACE] [-S FIRST_SOC] [-s] [-U] [-b file] [-c JSON] [-e]
+ [-f RAW] [-K] [-l file] [-O RAW] [-o VMEM] [-r ELF] [-w CSV]
+ [-x file] [-X] [-F TEST] [-k SECONDS] [-z] [-R] [-T FACTOR]
+ [-Z] [-v] [-V] [-d] [--log-time] [--debug LOGGER]
+ [--info LOGGER] [--warn LOGGER]
 
 OpenTitan QEMU unit test sequencer.
 
@@ -25,14 +26,18 @@ Virtual machine:
  per inst. or 'auto'
  -L LOG_FILE, --log_file LOG_FILE
  log file for trace and log messages
- -M LOG, --log LOG log message types
+ -M VARIANT, --variant VARIANT
+ machine variant (machine specific)
+ -N LOG, --log LOG log message types
  -m MACHINE, --machine MACHINE
  virtual machine (default to ot-earlgrey)
  -Q OPTS, --opts OPTS QEMU verbatim option (can be repeated)
  -q QEMU, --qemu QEMU path to qemu application (default: build/qemu-system-
  riscv32)
+ -P VCP, --vcp VCP serial port devices (default: use serial0)
  -p DEVICE, --device DEVICE
- serial port device name (default to localhost:8000)
+ serial port device name / template name (default to
+ localhost:8000)
  -t TRACE, --trace TRACE
  trace event definition file
  -S FIRST_SOC, --first-soc FIRST_SOC
@@ -73,6 +78,7 @@ Execution:
 
 Extras:
  -v, --verbose increase verbosity
+ -V, --vcp-verbose increase verbosity of QEMU virtual comm ports
  -d enable debug mode
  --log-time show local time in log messages
  --debug LOGGER assign debug level to logger(s)
@@ -95,12 +101,16 @@ This tool may be used in two ways, which can be combined:
  Use 'auto' to enable QEMU adaptive icount counter. Note that this option slows down the execution
  of guest applications.
 * `-L` / `--log_file` specify the log file for trace and log messages from QEMU.
-* `-M` / `--log` specify which log message types should be logged; most useful types are:
- * `in_asm` for guest instruction disassembly,
- * `unimp` for uimplemented guest features,
- * `int` for guest interrupts and exceptions,
- * `guest_errors` for unexpected guest behavior,
- * `exec` for guest execution stream (caution: highly verbose).
+* `-N` / `--log` specify which log message types should be logged; most useful types are:
+ * `in_asm`/`A` for guest instruction disassembly,
+ * `unimp`/`U` for uimplemented guest features,
+ * `int`/`I` for guest interrupts and exceptions,
+ * `guest_errors`/`G` for unexpected guest behavior,
+ * `exec`/`E` for guest execution stream (caution: highly verbose).
+ These definitions may be abbreviated using their upper case, single char variants, _e.g._
+ `-N GIU` would enable _guest_errors_, _interruptions_ and _unimplemented features_ logs.
+* `-M` / `--variant` specify a variant of the selected machine. The accepted values depend on the
+ machine, see `-m` option.
 * `-m` / `--machine` specify the kind of virtual machine to run.
 * `-q` / `--qemu` specify an alternative path to the QEMU application.
 * `-Q` / `--opts` add a single QEMU option forwarded verbatim to QEMU (no check is performed)
@@ -165,6 +175,7 @@ This tool may be used in two ways, which can be combined:
 
 ### Extras
 
+* `-V` / `--vcp-verbose` can be repeated to increase verbosity of the QEMU virtual comm ports
 * `-v` / `--verbose` can be repeated to increase verbosity of the script, mostly for debug purpose.
 * `-d` only useful to debug the script, reports any Python traceback to the standard error stream.
 * `--log-time` show local time before each logged message

docs/opentitan/spi_device.md

@@ -37,7 +37,6 @@ instanciated this way from the command line:
 
 ````
 -chardev socket,id=spidev,host=localhost,port=8004,server=on,wait=off
--global ot-spi_device.chardev=spidev
 ````
 
 Note that `opentitantool` and association library do support this protocol when the `qemu` backend
@@ -57,6 +56,9 @@ Commands:
  --qemu-spidev-port <QEMU_SPIDEV_PORT> [default: 8004]
 ````
 
+IT is also possible to use [`spidevflash.py`](spidevflash.md) tool to upload a binary using the same
+protocol.
+
 ### SPI device CharDev protocol
 
 SPI clock is not emulated, but each byte exchanged over the communication channel represent 8-bit

docs/opentitan/spidevflash.md

@@ -0,0 +1,47 @@
+# `spidevflash.py`
+
+`spidevflash.py` is a tiny script to upload bootstrap image using the SPI device virtual device.
+
+## Usage
+
+````text
+usage: spidevflash.py [-h] -f FILE [-a ADDRESS] [-r HOST] [-p PORT] [-v] [-d]
+
+SPI device flasher tool.
+
+options:
+ -h, --help show this help message and exit
+ -f FILE, --file FILE Binary file to flash
+ -a ADDRESS, --address ADDRESS
+ Address in the SPI flash (default to 0)
+ -r HOST, --host HOST remote host name (default: localhost)
+ -p PORT, --port PORT remote host TCP port (defaults to 8004)
+ -v, --verbose increase verbosity
+ -d, --debug enable debug mode
+````
+
+### Arguments
+
+* `-a` specify an alernative start address
+
+* `-d` only useful to debug the script, reports any Python traceback to the standard error stream.
+
+* `-f` specify the binary file to upload
+
+* `-p` specify an alternative port for the TCP connection on the QEMU instance
+
+* `-r` specify the name or address of the remote host running the QEMU instance
+
+* `-v` can be repeated to increase verbosity of the script, mostly for debug purpose.
+
+### Examples
+
+With the following examples:
+
+`-chardev socket,id=spidev,host=localhost,port=8004,server=on,wait=off -global ot-spi_device.chardev=spidev` has been
+added to the QEMU command line to create a TCP chardev and connect it the SPI Device backend. See the [SPI Device](spi_device.md) documentation for details.
+
+* Upload a bootstrap binary
+ ````sh
+ ./scripts/opentitan/spidevflash.py -f test_bootstrap_virtual_sim_dv+manifest.bin
+ ````

docs/opentitan/tools.md

@@ -44,6 +44,7 @@ directory to help with these tasks.
 * `ot-tidy.sh` is a simple shell wrapper to run clang-tidy (C linter) on OpenTitan files
 * `present.py` implements the Present 128-bit scrambler/descrambler used in OTP image files for
  HW digest verification.
+* [spidevice.py](spidevice.md) is a tiny script to upload a binary using the SPI device.
 * `treillis/` directory contains the test application to test the [GPIO](gpio.md) device.
 * [`uartmux.py`](uartmux.md) is a tiny stream wrapper to help dealing with multiple QEMU output
  streams, typically multiple virtual UARTs.

hw/opentitan/ot_csrng.c

@@ -1029,8 +1029,7 @@ ot_csrng_handle_instantiate(OtCSRNGState *s, unsigned slot)
  uint32_t command = ot_fifo32_peek(&inst->cmd_fifo);
  uint32_t clen = FIELD_EX32(command, OT_CSNRG_CMD, CLEN);
  bool flag0 =
- ot_csrng_check_multibitboot(s, FIELD_EX32(command, OT_CSNRG_CMD, FLAG0),
- ALERT_STATUS_BIT(FLAG0));
+ FIELD_EX32(command, OT_CSNRG_CMD, FLAG0) == OT_MULTIBITBOOL4_TRUE;
 
  uint32_t num;
  const uint32_t *buffer =
@@ -1124,8 +1123,7 @@ static OtCSRNDCmdResult ot_csrng_handle_reseed(OtCSRNGState *s, unsigned slot)
 
  uint32_t command = ot_fifo32_peek(&inst->cmd_fifo);
  bool flag0 =
- ot_csrng_check_multibitboot(s, FIELD_EX32(command, OT_CSNRG_CMD, FLAG0),
- ALERT_STATUS_BIT(FLAG0));
+ FIELD_EX32(command, OT_CSNRG_CMD, FLAG0) == OT_MULTIBITBOOL4_TRUE;
 
  xtrace_ot_csrng_info("reseed", flag0);