[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20210406090633.224f2921@gandalf.local.home>
Date: Tue, 6 Apr 2021 09:06:33 -0400
From: Steven Rostedt <rostedt@...dmis.org>
To: Zhang Jianhua <zhangjianhua18@...wei.com>
Cc: <mingo@...hat.com>, <linux-kernel@...r.kernel.org>,
<johnny.chenyi@...wei.com>, <heying24@...wei.com>
Subject: Re: [PATCH -next] arm: patch: FUNCTION_GRAPH_TRACER depends on
CPU_ENDIAN_BE8 or !CPU_ENDIAN_BE32
On Tue, 6 Apr 2021 18:23:51 +0800
Zhang Jianhua <zhangjianhua18@...wei.com> wrote:
> If CONFIG_FUNCTION_GRAPH_TRACER=y, the following errors will be seen
> while compiling patch.c
>
> arch/arm/kernel/patch.c: In function ‘__patch_text_real’:
> arch/arm/kernel/patch.c:94:11: error: implicit declaration of function
> ‘__opcode_to_mem_thumb32’; did you mean ‘__opcode_to_mem_thumb16’?
> [-Werror=implicit-function-declaration]
> insn = __opcode_to_mem_thumb32(insn);
> ^~~~~~~~~~~~~~~~~~~~~~~
> __opcode_to_mem_thumb16
>
> Signed-off-by: Zhang Jianhua <zhangjianhua18@...wei.com>
> ---
> kernel/trace/Kconfig | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
> index 7fa82778c3e6..fbf5fd80f105 100644
> --- a/kernel/trace/Kconfig
> +++ b/kernel/trace/Kconfig
> @@ -177,6 +177,7 @@ config FUNCTION_GRAPH_TRACER
> depends on HAVE_FUNCTION_GRAPH_TRACER
> depends on FUNCTION_TRACER
> depends on !X86_32 || !CC_OPTIMIZE_FOR_SIZE
> + depends on CPU_ENDIAN_BE8 || !CPU_ENDIAN_BE32
Doesn't this break 32 bit big endian powerpc?
Not to mention, my config doesn't have either of those.
NACK
-- Steve
> default y
> help
> Enable the kernel to trace a function at both its return
Powered by blists - more mailing lists