[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v2 1/3] hw/watchdog: wdt_ibex_aon.c: Implement the watchdog f
|
From: |
Thomas Huth |
|
Subject: |
Re: [PATCH v2 1/3] hw/watchdog: wdt_ibex_aon.c: Implement the watchdog for the OpenTitan |
|
Date: |
Tue, 27 Sep 2022 08:31:17 +0200 |
|
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Thunderbird/91.13.0 |
Hi Tyler!
On 27/09/2022 01.03, Tyler Ng wrote:
Hi Thomas,
On Thu, Sep 22, 2022 at 9:17 AM Thomas Huth <thuth@redhat.com
<mailto:thuth@redhat.com>> wrote:
On 22/09/2022 17.58, Tyler Ng wrote:
> This commit adds most of an implementation of the OpenTitan Always-On
> Timer. The documentation for this timer is found here:
>
> https://docs.opentitan.org/hw/ip/aon_timer/doc/
<https://docs.opentitan.org/hw/ip/aon_timer/doc/>
>
> Using commit 217a0168ba118503c166a9587819e3811eeb0c0c
>
> The implementation includes most of the watchdog features; it does not
> implement the wakeup timer.
>
> An important note: the OpenTitan board uses the sifive_plic. The plic
> will not be able to claim the bark interrupt (159) because the sifive
> plic sets priority[159], but checks priority[158] for the priority, so
> it thinks that the interrupt's priority is 0 (effectively disabled).
...
> diff --git a/tests/qtest/ibex-aon-timer-test.c
> b/tests/qtest/ibex-aon-timer-test.c
> new file mode 100644
> index 0000000000..af33feac39
> --- /dev/null
> +++ b/tests/qtest/ibex-aon-timer-test.c
> @@ -0,0 +1,189 @@
> +/*
> + * Testing the OpenTitan AON Timer
> + *
> + * Copyright (c) 2022 Rivos Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person
obtaining a copy
> + * of this software and associated documentation files (the
> "Software"), to deal
> + * in the Software without restriction, including without limitation
the rights
> + * to use, copy, modify, merge, publish, distribute, sublicense,
and/or sell
> + * copies of the Software, and to permit persons to whom the Software is
> + * furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be
included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT
SHALL
> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES
OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
> ARISING FROM,
> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
DEALINGS IN
> + * THE SOFTWARE.
Could you maybe add a SPDX license identifier at the beginning of the
comment, so that it's easier to identify the license at a first glance?
(also in the other new files)
Will do, was actually thinking of switching over to GPL-2.0-or-later as
opposed to MIT.
Yes, that would be the best fit for QEMU, I think.
> + */
> +
> +#include "qemu/osdep.h"
> +#include "libqtest.h"
> +#include "qapi/qmp/qdict.h"
> +
> +#define AON_BASE_ADDR (0x40470000ul)
> +#define AON_ADDR(addr) (AON_BASE_ADDR + addr)
> +#define AON_WKUP_IRQ 158
> +#define AON_BARK_IRQ 159
> +#define AON_FREQ 200000 /* 200 KHz */
> +#define AON_PERIOD_NS 5000
> +#define NANOS_PER_SECOND 1000000000LL
> +/* Test that reads work, and that the regs get reset to the correct
value */
> +static void test_reads(void)
> +{
> + QTestState *test = qtest_init("-M opentitan");
> + g_assert(qtest_readl(test, AON_ADDR(0x00)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x04)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x08)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x0c)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x10)) == 1); > +
g_assert(qtest_readl(test, AON_ADDR(0x14)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x18)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x1c)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x20)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x24)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x28)) == 0);
> + g_assert(qtest_readl(test, AON_ADDR(0x2c)) == 0);
The read tests that check for 0 could maybe be simplified with a for-loop
(or two).
I'm not entirely sure about what benefit this would bring after writing it out.
Mostly a matter of taste. Keep it in the current shape if you prefer that.
> + qtest_quit(test);
> +}
> +
> +static void test_writes(void)
> +{
> + /* Test that writes worked, while the config is unlocked */
> + QTestState *test = qtest_init("-M opentitan");
> +
> +
> + qtest_writel(test, AON_ADDR(0x18), (1 << 19)); /* WDOG_BARK_THOLD */
> + g_assert_cmpuint(qtest_readl(test, AON_ADDR(0x18)),
> + ==, (1 << 19));
> +
> + qtest_writel(test, AON_ADDR(0x1c), (1 << 20)); /* WDOG_BITE_THOLD */
> + g_assert_cmpuint(qtest_readl(test, AON_ADDR(0x1c)),
> + ==, (1 << 20));
> +
> + qtest_writel(test, AON_ADDR(0x14), 0x1); /* WDOG_CTRL enable */
> + g_assert_cmpuint(qtest_readl(test, AON_ADDR(0x14)),
> + ==, 0x1);
> +
> + qtest_writel(test, AON_ADDR(0x10), 0x0); /* WDOG_REGWEN enable */
> + g_assert_cmpuint(qtest_readl(test, AON_ADDR(0x10)), ==, 0x0);
I think the above code would be better readable if you'd provide a helper
function like this:
static void writel_and_assert(QTestState qts, int addr, int val)
{
qtest_writel(qts, AON_ADDR(addr), val);
g_assert_cmpuint(qtest_readl(qts, AON_ADDR(addr)), val);
}
Thanks for the suggestion. I decided to go with a macro instead though,
because it makes it easier to distinguish where an assertion failed without
a debugger.
That's a good idea, indeed!
Thomas