Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220708162711.1309633-1-dlatypov@google.com> <20220711144651.yekrivauz3phkuvm@houat>
In-Reply-To: <20220711144651.yekrivauz3phkuvm@houat>
From:   Daniel Latypov <dlatypov@google.com>
Date:   Mon, 11 Jul 2022 08:10:36 -0700
Message-ID: <CAGS_qxp=UAJ19zQOZKP8ywPiNAj4wT80ubH2YW060mmADoqBcw@mail.gmail.com>
Subject: Re: [PATCH v3] kunit: tool: Enable virtio/PCI by default on UML
To:     Maxime Ripard <maxime@cerno.tech>
Cc:     brendanhiggins@google.com, davidgow@google.com,
        linux-kernel@vger.kernel.org, kunit-dev@googlegroups.com,
        linux-kselftest@vger.kernel.org, skhan@linuxfoundation.org,
        =?UTF-8?B?Sm9zw6kgRXhww7NzaXRv?= <jose.exposito89@gmail.com>,
        Mike Turquette <mturquette@baylibre.com>,
        Stephen Boyd <sboyd@kernel.org>, linux-clk@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Jul 11, 2022 at 7:46 AM Maxime Ripard <maxime@cerno.tech> wrote:
>
> Unfortunately, this breaks the clock tests in next-20220711:
> $ ./tools/testing/kunit/kunit.py run --kunitconfig=drivers/clk/.kunitconfig  --raw_output

Thanks, this is indeed an issue.

I remember noticing this in early April.
I incorrectly remembered that a fix had been sent.

A more minimal reproducer:
$ ./tools/testing/kunit/kunit.py run --kunitconfig=drivers/clk
'clk-gate-test.clk_gate_test_enable'

The part of the test that becomes problematic with this patch (i.e.
enabling logic iomem) is the cast on line 143.

   130  struct clk_gate_test_context {
   131          void __iomem *fake_mem;
   132          struct clk_hw *hw;
   133          struct clk_hw *parent;
   134          u32 fake_reg; /* Keep at end, KASAN can detect out of bounds */
   135  };
   136
   137  static struct clk_gate_test_context
*clk_gate_test_alloc_ctx(struct kunit *test)
   138  {
   139          struct clk_gate_test_context *ctx;
   140
   141          test->priv = ctx = kunit_kzalloc(test, sizeof(*ctx),
GFP_KERNEL);
   142          KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ctx);
   143          ctx->fake_mem = (void __force __iomem *)&ctx->fake_reg;
   144
   145          return ctx;
   146  }

A simple fix we could carry in the KUnit branch is this:

diff --git a/drivers/clk/.kunitconfig b/drivers/clk/.kunitconfig
index cdbc7d7deba9..2fbeb71316f8 100644
--- a/drivers/clk/.kunitconfig
+++ b/drivers/clk/.kunitconfig
@@ -2,3 +2,4 @@ CONFIG_KUNIT=y
 CONFIG_COMMON_CLK=y
 CONFIG_CLK_KUNIT_TEST=y
 CONFIG_CLK_GATE_KUNIT_TEST=y
+CONFIG_UML_PCI_OVER_VIRTIO=n

The new ability to disable it comes from
https://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest.git/commit/?h=kunit&id=8a7c6f859a20ca36a9e3ce71662de697898c9ef5

Thoughts?

Daniel