[PATCH] arm64: cache: Lower ARCH_DMA_MINALIGN to 64 (L1_CACHE_BYTES)

[Robin Murphy]

On 2021-07-06 11:26, Catalin Marinas wrote:
> On Tue, Jul 06, 2021 at 09:26:59AM +0000, Yassine Oudjana wrote:
>> In-Reply-To: <20210527124356.22367-1-will at kernel.org>
>>> Reduce ARCH_DMA_MINALIGN to 64 bytes and allow the warning/taint to
>>> indicate if there are machines that unknowingly rely on this.
>>
>> The warning is being triggered on Qualcomm MSM8996, as well as the out-of-spec taint:
> 
> Is this booting with ACPI or DT?
> 
>> ------------[ cut here ]------------
>> rtc-pm8xxx 400f000.qcom,spmi:pmic at 0:rtc at 6000: ARCH_DMA_MINALIGN smaller than CTR_EL0.CWG (64 < 128)
>> WARNING: CPU: 0 PID: 1 at arch/arm64/mm/dma-mapping.c:45 arch_setup_dma_ops+0xf8/0x10c
> [...]
>> This warning is triggered with nearly every driver probe, not only rtc-pm8xxx.
> 
> I have a suspicion none of the reported devices actually do any DMA, so
> in practice it should be safe but we need to figure out why
> arch_setup_dma_ops() gets called.

It gets called because there's no straightforward way to know that a 
platform device *isn't* DMA-capable, so we have to assume they are.

I would also assume that in a Qcom SoC there really are at least some 
things doing non-coherent DMA :(

Robin.