Re: [PATCHv5 3/4] tracing: Add register read/write tracing support

From: Sai Prakash Ranjan <quic_saipraka@quicinc.com>
To: Arnd Bergmann <arnd@arndb.de>
Cc: Will Deacon <will@kernel.org>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Steven Rostedt <rostedt@goodmis.org>,
	Marc Zyngier <maz@kernel.org>,
	gregkh <gregkh@linuxfoundation.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Linux ARM <linux-arm-kernel@lists.infradead.org>,
	linux-arm-msm <linux-arm-msm@vger.kernel.org>,
	<quic_psodagud@quicinc.com>,
	Prasad Sodagudi <psodagud@codeaurora.org>
Subject: Re: [PATCHv5 3/4] tracing: Add register read/write tracing support
Date: Mon, 6 Dec 2021 15:43:50 +0530	[thread overview]
Message-ID: <429d1354-f128-205a-ae27-3eadd8929209@quicinc.com> (raw)
In-Reply-To: <CAK8P3a1wOEkoteJRu+qpE0k3gJ0fRnwfn1Zrtcnmgn5wT4yQdw@mail.gmail.com>

On 12/6/2021 2:29 PM, Arnd Bergmann wrote:
> On Mon, Dec 6, 2021 at 9:28 AM Sai Prakash Ranjan
> <quic_saipraka@quicinc.com> wrote:
>> diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
>> index 420ff4bc67fd..9f55bcc51de1 100644
>> --- a/kernel/trace/Kconfig
>> +++ b/kernel/trace/Kconfig
>> @@ -95,6 +95,13 @@ config RING_BUFFER_ALLOW_SWAP
>>           Allow the use of ring_buffer_swap_cpu.
>>           Adds a very slight overhead to tracing when enabled.
>>
>> +config TRACE_MMIO_ACCESS
>> +       bool "Register read/write tracing"
>> +       depends on TRACING
>> +       help
>> +         Create tracepoints for MMIO read/write operations. These trace events
>> +         can be used for logging all MMIO read/write operations.
> I think this needs a 'depends on ARCH_HAVE_TRACE_MMIO_ACCESS'
> or similar.

Sure, will add it.

>> +void log_read_mmio(u8 width, const volatile void __iomem *addr)
>> +{
>> +       trace_rwmmio_read(CALLER_ADDR0, CALLER_ADDR1, width, addr);
>> +}
> Here, it may be better to pass the caller address as an argument, I think
> CALLER_ADDR1 is not always reliable, though it's possible that it is
> in the configurations when this file gets enabled.
>
>

Do you mean that we use __builtin_return_address(0,1) directly here or 
that I pass
__func__ as the argument to log_read/write_mmio or is there some other 
way to
pass the caller address?

Thanks,
Sai