[PATCH] riscv: timer: Add support for an early timer

Rick Chen rickchen36 at gmail.com
Fri Nov 27 08:58:30 CET 2020


Hi, Pragnesh

> Hi Rick,
>
> >-----Original Message-----
> >From: Rick Chen <rickchen36 at gmail.com>
> >Sent: 26 November 2020 14:44
> >To: Pragnesh Patel <pragnesh.patel at openfive.com>
> >Cc: Simon Glass <sjg at chromium.org>; U-Boot Mailing List <u-
> >boot at lists.denx.de>; Atish Patra <atish.patra at wdc.com>; Bin Meng
> ><bmeng.cn at gmail.com>; Paul Walmsley ( Sifive) <paul.walmsley at sifive.com>;
> >Anup Patel <anup.patel at wdc.com>; Sagar Kadam
> ><sagar.kadam at openfive.com>; Palmer Dabbelt <palmer at dabbelt.com>; rick
> ><rick at andestech.com>; Alan Kao <alankao at andestech.com>; Leo Liang
> ><ycliang at andestech.com>
> >Subject: Re: [PATCH] riscv: timer: Add support for an early timer
> >
> >[External Email] Do not click links or attachments unless you recognize the
> >sender and know the content is safe
> >
> >Hi Pragnesh
> >
> >> Hi Rick,
> >>
> >> >-----Original Message-----
> >> >From: Rick Chen <rickchen36 at gmail.com>
> >> >Sent: 24 November 2020 13:08
> >> >To: Pragnesh Patel <pragnesh.patel at openfive.com>
> >> >Cc: U-Boot Mailing List <u-boot at lists.denx.de>; Atish Patra
> >> ><atish.patra at wdc.com>; Bin Meng <bmeng.cn at gmail.com>; Paul Walmsley (
> >> >Sifive) <paul.walmsley at sifive.com>; Anup Patel <anup.patel at wdc.com>;
> >> >Sagar Kadam <sagar.kadam at openfive.com>; Palmer Dabbelt
> >> ><palmer at dabbelt.com>; Simon Glass <sjg at chromium.org>; rick
> >> ><rick at andestech.com>; Alan Kao <alankao at andestech.com>; Leo Liang
> >> ><ycliang at andestech.com>
> >> >Subject: Re: [PATCH] riscv: timer: Add support for an early timer
> >> >
> >> >[External Email] Do not click links or attachments unless you
> >> >recognize the sender and know the content is safe
> >> >
> >> >Hi Pragnesh,
> >> >
> >> >> From: Pragnesh Patel [mailto:pragnesh.patel at sifive.com]
> >> >> Sent: Tuesday, November 17, 2020 7:05 PM
> >> >> To: u-boot at lists.denx.de
> >> >> Cc: atish.patra at wdc.com; palmerdabbelt at google.com;
> >> >bmeng.cn at gmail.com;
> >> >> paul.walmsley at sifive.com; anup.patel at wdc.com;
> >> >> sagar.kadam at sifive.com; Rick Jian-Zhi Chen(陳建志); Pragnesh Patel;
> >> >> Palmer Dabbelt; Sean Anderson; Simon Glass; Bin Meng
> >> >> Subject: [PATCH] riscv: timer: Add support for an early timer
> >> >>
> >> >> Added support for timer_early_get_count() and
> >> >> timer_early_get_rate() This is mostly useful in tracing.
> >> >>
> >> >> Signed-off-by: Pragnesh Patel <pragnesh.patel at sifive.com>
> >> >> ---
> >> >>  drivers/timer/andes_plmt_timer.c   | 21 ++++++++++++++++++++-
> >> >>  drivers/timer/riscv_timer.c        | 21 ++++++++++++++++++++-
> >> >>  drivers/timer/sifive_clint_timer.c | 21 ++++++++++++++++++++-
> >> >>  include/configs/ax25-ae350.h       |  5 +++++
> >> >>  include/configs/sifive-fu540.h     |  5 +++++
> >> >>  5 files changed, 70 insertions(+), 3 deletions(-)
> >> >>
> >> >
> >> >I verify with ae350_rv64_defconfig
> >> >
> >> >make FTRACE=1 ae350_rv64_defconfig
> >> >make FTRACE=1
> >> >
> >> >and it boot fail as below:
> >> >
> >> >U-Boot 2021.01-rc2-00140-geb42715 (Nov 24 2020 - 15:02:18 +0800)
> >> >
> >> >DRAM:  1 GiB
> >> >trace: enabled
> >> >
> >> >DO you have any suggestions ?
> >>
> >> Please enable CONFIG_TIMER_EARLY=y in ae350_rv64_defconfig
> >>
> >> Actually in v2, I will make TRACE to select TIMER_EARLY like below,
> >>
> >> --- a/lib/Kconfig
> >> +++ b/lib/Kconfig
> >> @@ -210,6 +210,7 @@ config BITREVERSE
> >>  config TRACE
> >>         bool "Support for tracing of function calls and timing"
> >>         imply CMD_TRACE
> >> +       select TIMER_EARLY
> >>
> >> Let me know if you have any suggestion.
> >
> >OK.
> >
> >After add CONFIG_TIMER_EARLY, U-Boot boots ok.
> >But When I try to booting kernel with FTRACE=1, following are the test stats:
> >
> >ae350_rv64_spl_defconfig without FTRACE=1, kernel booting is ok.
> >ae350_rv64_spl_defconfig with FTRACE=1, kernel booting fail.
> >ae350_rv64_defconfig with FTRACE=1, kernel booting is ok
> >
> >The failure case seems not reasonable.
> >Any suggestions ?
>
> Strange, Can you please tell me which steps you follow and also send some debug logs  if possible.
>

Following are the configurations, steps and debug logs:

+++ b/configs/ae350_rv64_defconfig
q+CONFIG_TRACE=y
+CONFIG_TRACE_BUFFER_SIZE=0x01000000
+CONFIG_TRACE_CALL_DEPTH_LIMIT=15
+CONFIG_CMD_TRACE=y
+CONFIG_TIMER_EARLY=y

+++ b/configs/ae350_rv64_spl_defconfig
+CONFIG_TRACE=y
+CONFIG_TRACE_BUFFER_SIZE=0x01000000
+CONFIG_TRACE_CALL_DEPTH_LIMIT=15
+CONFIG_CMD_TRACE=y
+CONFIG_TIMER_EARLY=y

//////////////////////////////////////////////// case 1
///////////////////////////////////////////////////
ae350_rv64_defconfig with FTRACE=1, kernel booting is ok
//////////////////////////////////////////////// case 1
///////////////////////////////////////////////////
make FTRACE=1 ae350_rv64_defconfig
make FTRACE=1
///////////////////////////////////////////////////////////////////////////////////////////////////////////////
U-Boot 2021.01-rc2-00139-gb3d3d69-dirty (Nov 25 2020 - 11:14:28 +0800)

DRAM:  1 GiB
trace: enabled
Flash: 64 MiB
MMC:   mmc at f0e00000: 0
Loading Environment from SPIFlash... SF: Detected mx25u1635e with page
size 256 Bytes, erase size 4 KiB, total 2 MiB
OK
In:    serial at f0300000
Out:   serial at f0300000
Err:   serial at f0300000
Net:   no alias for ethernet0
Warning: mac at e0100000 (eth0) using random MAC address - de:fa:3e:1b:11:42
eth0: mac at e0100000
Hit any key to stop autoboot:  0
RISC-V # fatload mmc 0:1 0x20000000 ae350_rv64_smp_4_no_fd_coherent.dtb
6455 bytes read in 67 ms (93.8 KiB/s)
RISC-V # fatload mmc 0:1 0x00600000 bootm_ae350_rv64_smp_bbl.bin
22518836 bytes read in 11915 ms (1.8 MiB/s)
RISC-V # bootm 0x00600000 - 0x20000000
## Booting kernel from Legacy Image at 00600000 ...
   Image Name:
   Image Type:   RISC-V Linux Kernel Image (uncompressed)
   Data Size:    22518772 Bytes = 21.5 MiB
   Load Address: 00000000
   Entry Point:  00000000
   Verifying Checksum ... OK
## Flattened Device Tree blob at 20000000
   Booting using the fdt blob at 0x20000000
   Loading Kernel Image
   Loading Device Tree to 000000001effb000, end 000000001efff936 ... OK

Starting kernel ...(fake run for tracing)

Starting kernel ...

OF: fdt: Ignoring memory range 0x0 - 0x200000
Linux version 4.17.0-00253-g49136e10bcb2 (sqa at atcsqa07) (gcc version
7.3.0 (2019-04-06_nds64le-linux-glibc-v5_experimental)) #1 SMP PREEMPT
Sat Apr 6 23:41:49 CST 2019
bootconsole [early0] enabled
Initial ramdisk at: 0x        (ptrval) (13665712 bytes)
Zone ranges:
  DMA32    [mem 0x0000000000200000-0x000000003fffffff]
  Normal   empty
Movable zone start for each node
Early memory node ranges
...
...

//////////////////////////////////////////////// case 2
///////////////////////////////////////////////////
ae350_rv64_spl_defconfig with FTRACE=1, kernel booting fail
//////////////////////////////////////////////// case 2
///////////////////////////////////////////////////
make FTRACE=1 ae350_rv64_spl_defconfig
make FTRACE=1
///////////////////////////////////////////////////////////////////////////////////////////////////////////////
U-Boot SPL 2020.10-rc2-00175-gfa50824 (Sep 15 2020 - 19:26:29 +0800)
Trying to boot from MMC1
U-Boot SPL 2021.01-rc2-00139-gb3d3d69-dirty (Nov 25 2020 - 13:48:05 +0800)
Trying to boot from RAM
U-Boot 2021.01-rc2-00139-gb3d3d69-dirty (Nov 25 2020 - 13:48:05 +0800)
DRAM:  1 GiB
trace: enabled
Flash: 64 MiB
MMC:   mmc at f0e00000: 0
Loading Environment from SPIFlash... SF: Detected mx25u1635e with page
size 256 Bytes, erase size 4 KiB, total 2 MiB
OK
In:    serial at f0300000
Out:   serial at f0300000
Err:   serial at f0300000
Net:   no alias for ethernet0
Warning: mac at e0100000 (eth0) using random MAC address - 36:86:da:0f:8e:8d
eth0: mac at e0100000
Hit any key to stop autoboot:  0
27689996 bytes read in 15024 ms (1.8 MiB/s)
6435 bytes read in 25 ms (251 KiB/s)
## Booting kernel from Legacy Image at 00600000 ...
   Image Name:
   Image Type:   RISC-V Linux Kernel Image (uncompressed)
   Data Size:    27689932 Bytes = 26.4 MiB
   Load Address: 00200000
   Entry Point:  00200000
   Verifying Checksum ... OK
## Flattened Device Tree blob at 20000000
   Booting using the fdt blob at 0x20000000
   Loading Kernel Image
   Loading Device Tree to 000000001effb000, end 000000001efff922 ... OK

Starting kernel ...(fake run for tracing)

Starting kernel ...

(hang here)

//////////////////////////////////////////////// case 3
///////////////////////////////////////////////////
ae350_rv64_spl_defconfig without FTRACE=1, kernel booting ok
//////////////////////////////////////////////// case 2
///////////////////////////////////////////////////
make ae350_rv64_spl_defconfig
make
///////////////////////////////////////////////////////////////////////////////////////////////////////////////

U-Boot SPL 2021.01-rc2-00139-gb3d3d69-dirty (Nov 25 2020 - 11:25:30 +0800)
Trying to boot from RAM
U-Boot 2021.01-rc2-00139-gb3d3d69-dirty (Nov 25 2020 - 11:25:30 +0800)
DRAM:  1 GiB
trace: enabled
Flash: 64 MiB
MMC:   mmc at f0e00000: 0
Loading Environment from SPIFlash... SF: Detected mx25u1635e with page
size 256 Bytes, erase size 4 KiB, total 2 MiB
OK
In:    serial at f0300000
Out:   serial at f0300000
Err:   serial at f0300000
Net:   no alias for ethernet0
Warning: mac at e0100000 (eth0) using random MAC address - 6a:1a:a0:e2:cc:d8
eth0: mac at e0100000
Hit any key to stop autoboot:  0
27689996 bytes read in 9333 ms (2.8 MiB/s)
6435 bytes read in 15 ms (418.9 KiB/s)
## Booting kernel from Legacy Image at 00600000 ...
   Image Name:
   Image Type:   RISC-V Linux Kernel Image (uncompressed)
   Data Size:    27689932 Bytes = 26.4 MiB
   Load Address: 00200000
   Entry Point:  00200000
   Verifying Checksum ... OK
## Flattened Device Tree blob at 20000000
   Booting using the fdt blob at 0x20000000
   Loading Kernel Image
   Loading Device Tree to 000000001effb000, end 000000001efff922 ... OK

Starting kernel ...(fake run for tracing)

Starting kernel ...

[    0.000000] OF: fdt: Ignoring memory range 0x0 - 0x200000
[    0.000000] Linux version 5.4.0-00137-g5601822 (rick at atcsqa06) (gcc
version 7.4.0 (2020-07-29_nds64le-linux-glibc-v5d-532904c2315e)) #56
SMP PREEMPT Thu Nov 19 18:06:46 CST 2020
[    0.000000] initrd not found or empty - disabling initrd
[    0.000000] Zone ranges:
[    0.000000]   DMA32    [mem 0x0000000000200000-0x000000003fffffff]
[    0.000000]   Normal   empty
[    0.000000] Movable zone start for each node
[    0.000000] Early memory node ranges
...
////////////////////////////////////////////////// end
//////////////////////////////////////////////////////////

Thanks,
Rick

> >
> >Thanks,
> >Rick
> >
> >>
> >> >
> >> >Thanks,
> >> >Rick
> >> >
> >> >> diff --git a/drivers/timer/andes_plmt_timer.c
> >> >> b/drivers/timer/andes_plmt_timer.c
> >> >> index cec86718c7..74b795c97a 100644
> >> >> --- a/drivers/timer/andes_plmt_timer.c
> >> >> +++ b/drivers/timer/andes_plmt_timer.c
> >> >> @@ -17,11 +17,30 @@
> >> >>  /* mtime register */
> >> >>  #define MTIME_REG(base)                        ((ulong)(base))
> >> >>
> >> >> -static u64 andes_plmt_get_count(struct udevice *dev)
> >> >> +static u64 notrace andes_plmt_get_count(struct udevice *dev)
> >> >>  {
> >> >>         return readq((void __iomem *)MTIME_REG(dev->priv));  }
> >> >>
> >> >> +#if CONFIG_IS_ENABLED(RISCV_MMODE)
> >> >> +/**
> >> >> + * timer_early_get_rate() - Get the timer rate before driver model
> >> >> +*/ unsigned long notrace timer_early_get_rate(void) {
> >> >> +       return RISCV_MMODE_TIMER_FREQ; }
> >> >> +
> >> >> +/**
> >> >> + * timer_early_get_count() - Get the timer count before driver
> >> >> +model
> >> >> + *
> >> >> + */
> >> >> +u64 notrace timer_early_get_count(void) {
> >> >> +       return readq((void __iomem
> >> >> +*)MTIME_REG(RISCV_MMODE_TIMERBASE));
> >> >> +}
> >> >> +#endif
> >> >> +
> >> >>  static const struct timer_ops andes_plmt_ops = {
> >> >>         .get_count = andes_plmt_get_count,  }; diff --git
> >> >> a/drivers/timer/riscv_timer.c b/drivers/timer/riscv_timer.c index
> >> >> 21ae184057..a0f71ca897 100644
> >> >> --- a/drivers/timer/riscv_timer.c
> >> >> +++ b/drivers/timer/riscv_timer.c
> >> >> @@ -16,7 +16,7 @@
> >> >>  #include <timer.h>
> >> >>  #include <asm/csr.h>
> >> >>
> >> >> -static u64 riscv_timer_get_count(struct udevice *dev)
> >> >> +static u64 notrace riscv_timer_get_count(struct udevice *dev)
> >> >>  {
> >> >>         __maybe_unused u32 hi, lo;
> >> >>
> >> >> @@ -31,6 +31,25 @@ static u64 riscv_timer_get_count(struct udevice
> >*dev)
> >> >>         return ((u64)hi << 32) | lo;  }
> >> >>
> >> >> +#if CONFIG_IS_ENABLED(RISCV_SMODE)
> >> >> +/**
> >> >> + * timer_early_get_rate() - Get the timer rate before driver model
> >> >> +*/ unsigned long notrace timer_early_get_rate(void) {
> >> >> +       return RISCV_SMODE_TIMER_FREQ; }
> >> >> +
> >> >> +/**
> >> >> + * timer_early_get_count() - Get the timer count before driver
> >> >> +model
> >> >> + *
> >> >> + */
> >> >> +u64 notrace timer_early_get_count(void) {
> >> >> +       return riscv_timer_get_count(NULL); } #endif
> >> >> +
> >> >>  static int riscv_timer_probe(struct udevice *dev)  {
> >> >>         struct timer_dev_priv *uc_priv = dev_get_uclass_priv(dev);
> >> >> diff --git a/drivers/timer/sifive_clint_timer.c
> >> >> b/drivers/timer/sifive_clint_timer.c
> >> >> index 00ce0f08d6..9ae05a0e7e 100644
> >> >> --- a/drivers/timer/sifive_clint_timer.c
> >> >> +++ b/drivers/timer/sifive_clint_timer.c
> >> >> @@ -14,11 +14,30 @@
> >> >>  /* mtime register */
> >> >>  #define MTIME_REG(base)                        ((ulong)(base) + 0xbff8)
> >> >>
> >> >> -static u64 sifive_clint_get_count(struct udevice *dev)
> >> >> +static u64 notrace sifive_clint_get_count(struct udevice *dev)
> >> >>  {
> >> >>         return readq((void __iomem *)MTIME_REG(dev->priv));  }
> >> >>
> >> >> +#if CONFIG_IS_ENABLED(RISCV_MMODE)
> >> >> +/**
> >> >> + * timer_early_get_rate() - Get the timer rate before driver model
> >> >> +*/ unsigned long notrace timer_early_get_rate(void) {
> >> >> +       return RISCV_MMODE_TIMER_FREQ; }
> >> >> +
> >> >> +/**
> >> >> + * timer_early_get_count() - Get the timer count before driver
> >> >> +model
> >> >> + *
> >> >> + */
> >> >> +u64 notrace timer_early_get_count(void) {
> >> >> +       return readq((void __iomem
> >> >> +*)MTIME_REG(RISCV_MMODE_TIMERBASE));
> >> >> +}
> >> >> +#endif
> >> >> +
> >> >>  static const struct timer_ops sifive_clint_ops = {
> >> >>         .get_count = sifive_clint_get_count,  }; diff --git
> >> >> a/include/configs/ax25-ae350.h b/include/configs/ax25-ae350.h index
> >> >> b2606e794d..bd9c371f83 100644
> >> >> --- a/include/configs/ax25-ae350.h
> >> >> +++ b/include/configs/ax25-ae350.h
> >> >> @@ -17,6 +17,11 @@
> >> >>  #endif
> >> >>  #endif
> >> >>
> >> >> +#define RISCV_MMODE_TIMERBASE           0xe6000000
> >> >> +#define RISCV_MMODE_TIMER_FREQ          60000000
> >> >> +
> >> >> +#define RISCV_SMODE_TIMER_FREQ          60000000
> >> >> +
> >> >>  /*
> >> >>   * CPU and Board Configuration Options
> >> >>   */
> >> >> diff --git a/include/configs/sifive-fu540.h
> >> >> b/include/configs/sifive-fu540.h index c1c79db147..0d69d1c548
> >> >> 100644
> >> >> --- a/include/configs/sifive-fu540.h
> >> >> +++ b/include/configs/sifive-fu540.h
> >> >> @@ -36,6 +36,11 @@
> >> >>
> >> >>  #define CONFIG_STANDALONE_LOAD_ADDR    0x80200000
> >> >>
> >> >> +#define RISCV_MMODE_TIMERBASE          0x2000000
> >> >> +#define RISCV_MMODE_TIMER_FREQ         1000000
> >> >> +
> >> >> +#define RISCV_SMODE_TIMER_FREQ         1000000
> >> >> +
> >> >>  /* Environment options */
> >> >>
> >> >>  #ifndef CONFIG_SPL_BUILD
> >> >> --
> >> >> 2.17.1


More information about the U-Boot mailing list