[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <D3ZX6LZNLOYC.30CT6AX32VI2T@cknow.org>
Date: Sat, 07 Sep 2024 10:47:18 +0200
From: "Diederik de Haas" <didi.debian@...ow.org>
To: "Min-Hua Chen" <minhuadotchen@...il.com>
Cc: <airlied@...il.com>, <andy.yan@...k-chips.com>, <andyshrk@....com>,
<daniel@...ll.ch>, <dri-devel@...ts.freedesktop.org>, <heiko@...ech.de>,
<hjc@...k-chips.com>, <linux-arm-kernel@...ts.infradead.org>,
<linux-kernel@...r.kernel.org>, <linux-rockchip@...ts.infradead.org>,
<maarten.lankhorst@...ux.intel.com>, <mripard@...nel.org>,
<tzimmermann@...e.de>
Subject: Re: [PATCH] drm/rockchip: include rockchip_drm_drv.h
On Sat Sep 7, 2024 at 5:02 AM CEST, Min-Hua Chen wrote:
> >FWIW: I didn't see it either, but I assumed I was missing the right
> >context (i.e. patches) needed to trigger that warning.
>
> I triggered the warning by the following step:
>
> install 'sparse' first
>
> ARCH=arm64 LLVM=1 make C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' mrproper defconfig all -j8
This, especially the 'LLVM' part, is important context information
and should be part of the commit message.
I had only just started when I saw a number of sparse warnings:
DTC arch/arm64/boot/dts/rockchip/rk3588-edgeble-neu6a-io.dtb
OVL arch/arm64/boot/dts/ti/k3-j721e-evm.dtb
OVL arch/arm64/boot/dts/ti/k3-j721s2-evm.dtb
OVL arch/arm64/boot/dts/ti/k3-am654-idk.dtb
DTC arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r3-kb.dtb
DTC arch/arm64/boot/dts/rockchip/rk3588-edgeble-neu6a-wifi.dtbo
DTC arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r3-lte.dtb
DTC arch/arm64/boot/dts/rockchip/rk3588-edgeble-neu6b-io.dtb
DTC arch/arm64/boot/dts/rockchip/rk3588-evb1-v10.dtb
DTC arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r9.dtb
../init/main.c:192:12: sparse: warning: symbol 'envp_init' was not declared. Should it be static?
../init/main.c:290:16: sparse: warning: cast to restricted __le32
../init/main.c:291:16: sparse: warning: cast to restricted __le32
CHECK ../init/do_mounts.c
And several followed, including in c-code files. So I stopped the build
and assume you've identified a or several actual issues.
I've seen several commits where changes were made because LLVM flagged
potentially problematic code, where GCC did not, so it's quite possible
you're on to something here.
But it would be helpful if the commit message said what code was
potentially problematic and why. And then the proper fix for that could
indeed be to include `rockchip_drm_drv.h`.
Cheers,
Diederik
Download attachment "signature.asc" of type "application/pgp-signature" (229 bytes)
Powered by blists - more mailing lists