* [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable
@ 2022-04-09 15:35 Chengming Zhou
2022-04-09 15:35 ` [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly Chengming Zhou
` (2 more replies)
0 siblings, 3 replies; 7+ messages in thread
From: Chengming Zhou @ 2022-04-09 15:35 UTC (permalink / raw)
To: rostedt, mingo, catalin.marinas, will, tglx, bp, dave.hansen,
x86, hpa, broonie, mark.rutland, ardb, zhouchengming,
linux-arm-kernel
Cc: linux-kernel, duanxiongchun, songmuchun
The ftrace_[enable,disable]_ftrace_graph_caller() are used to do
special hooks for graph tracer, which are not needed on some ARCHs
that use graph_ops:func function to install return_hooker.
So introduce the weak version in ftrace core code to cleanup
in x86.
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
---
v4:
- put weak ftrace_enable,disable_ftrace_graph_caller() in
fgraph.c instead of ftrace.c as suggested by Steve.
v3:
- consolidate two #if into a single #if, suggested by Steve. Thanks.
---
arch/x86/kernel/ftrace.c | 17 ++---------------
kernel/trace/fgraph.c | 18 ++++++++++++++++++
2 files changed, 20 insertions(+), 15 deletions(-)
diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
index 1e31c7d21597..b09d73c2ba89 100644
--- a/arch/x86/kernel/ftrace.c
+++ b/arch/x86/kernel/ftrace.c
@@ -579,9 +579,7 @@ void arch_ftrace_trampoline_free(struct ftrace_ops *ops)
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
-#ifdef CONFIG_DYNAMIC_FTRACE
-
-#ifndef CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS
+#if defined(CONFIG_DYNAMIC_FTRACE) && !defined(CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS)
extern void ftrace_graph_call(void);
static const char *ftrace_jmp_replace(unsigned long ip, unsigned long addr)
{
@@ -610,18 +608,7 @@ int ftrace_disable_ftrace_graph_caller(void)
return ftrace_mod_jmp(ip, &ftrace_stub);
}
-#else /* !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
-int ftrace_enable_ftrace_graph_caller(void)
-{
- return 0;
-}
-
-int ftrace_disable_ftrace_graph_caller(void)
-{
- return 0;
-}
-#endif /* CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
-#endif /* !CONFIG_DYNAMIC_FTRACE */
+#endif /* CONFIG_DYNAMIC_FTRACE && !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
/*
* Hook the return address and push it in the stack of return addrs
diff --git a/kernel/trace/fgraph.c b/kernel/trace/fgraph.c
index 8f4fb328133a..289311680c29 100644
--- a/kernel/trace/fgraph.c
+++ b/kernel/trace/fgraph.c
@@ -30,6 +30,24 @@ int ftrace_graph_active;
/* Both enabled by default (can be cleared by function_graph tracer flags */
static bool fgraph_sleep_time = true;
+/*
+ * archs can override this function if they must do something
+ * to enable hook for graph tracer.
+ */
+int __weak ftrace_enable_ftrace_graph_caller(void)
+{
+ return 0;
+}
+
+/*
+ * archs can override this function if they must do something
+ * to disable hook for graph tracer.
+ */
+int __weak ftrace_disable_ftrace_graph_caller(void)
+{
+ return 0;
+}
+
/**
* ftrace_graph_stop - set to permanently disable function graph tracing
*
--
2.35.1
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly
2022-04-09 15:35 [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable Chengming Zhou
@ 2022-04-09 15:35 ` Chengming Zhou
2022-04-19 12:55 ` Mark Rutland
2022-04-18 13:24 ` [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable Chengming Zhou
2022-04-19 11:41 ` Mark Rutland
2 siblings, 1 reply; 7+ messages in thread
From: Chengming Zhou @ 2022-04-09 15:35 UTC (permalink / raw)
To: rostedt, mingo, catalin.marinas, will, tglx, bp, dave.hansen,
x86, hpa, broonie, mark.rutland, ardb, zhouchengming,
linux-arm-kernel
Cc: linux-kernel, duanxiongchun, songmuchun
As we do in commit 0c0593b45c9b ("x86/ftrace: Make function graph
use ftrace directly"), we don't need special hook for graph tracer,
but instead we use graph_ops:func function to install return_hooker.
Since commit 3b23e4991fb6 ("arm64: implement ftrace with regs") add
implementation for FTRACE_WITH_REGS on arm64, we can easily adopt
the same cleanup on arm64.
And this cleanup only changes the FTRACE_WITH_REGS implementation,
so the mcount-based implementation is unaffected.
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
---
v3:
- Add comments in ftrace_graph_func() as suggested by Steve. Thanks.
v2:
- Remove FTRACE_WITH_REGS ftrace_graph_caller asm, thanks Mark.
---
arch/arm64/include/asm/ftrace.h | 7 +++++++
arch/arm64/kernel/entry-ftrace.S | 17 -----------------
arch/arm64/kernel/ftrace.c | 17 +++++++++++++++++
3 files changed, 24 insertions(+), 17 deletions(-)
diff --git a/arch/arm64/include/asm/ftrace.h b/arch/arm64/include/asm/ftrace.h
index 1494cfa8639b..dbc45a4157fa 100644
--- a/arch/arm64/include/asm/ftrace.h
+++ b/arch/arm64/include/asm/ftrace.h
@@ -80,8 +80,15 @@ static inline unsigned long ftrace_call_adjust(unsigned long addr)
#ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
struct dyn_ftrace;
+struct ftrace_ops;
+struct ftrace_regs;
+
int ftrace_init_nop(struct module *mod, struct dyn_ftrace *rec);
#define ftrace_init_nop ftrace_init_nop
+
+void ftrace_graph_func(unsigned long ip, unsigned long parent_ip,
+ struct ftrace_ops *op, struct ftrace_regs *fregs);
+#define ftrace_graph_func ftrace_graph_func
#endif
#define ftrace_return_address(n) return_address(n)
diff --git a/arch/arm64/kernel/entry-ftrace.S b/arch/arm64/kernel/entry-ftrace.S
index e535480a4069..d42a205ef625 100644
--- a/arch/arm64/kernel/entry-ftrace.S
+++ b/arch/arm64/kernel/entry-ftrace.S
@@ -97,12 +97,6 @@ SYM_CODE_START(ftrace_common)
SYM_INNER_LABEL(ftrace_call, SYM_L_GLOBAL)
bl ftrace_stub
-#ifdef CONFIG_FUNCTION_GRAPH_TRACER
-SYM_INNER_LABEL(ftrace_graph_call, SYM_L_GLOBAL) // ftrace_graph_caller();
- nop // If enabled, this will be replaced
- // "b ftrace_graph_caller"
-#endif
-
/*
* At the callsite x0-x8 and x19-x30 were live. Any C code will have preserved
* x19-x29 per the AAPCS, and we created frame records upon entry, so we need
@@ -127,17 +121,6 @@ ftrace_common_return:
ret x9
SYM_CODE_END(ftrace_common)
-#ifdef CONFIG_FUNCTION_GRAPH_TRACER
-SYM_CODE_START(ftrace_graph_caller)
- ldr x0, [sp, #S_PC]
- sub x0, x0, #AARCH64_INSN_SIZE // ip (callsite's BL insn)
- add x1, sp, #S_LR // parent_ip (callsite's LR)
- ldr x2, [sp, #PT_REGS_SIZE] // parent fp (callsite's FP)
- bl prepare_ftrace_return
- b ftrace_common_return
-SYM_CODE_END(ftrace_graph_caller)
-#endif
-
#else /* CONFIG_DYNAMIC_FTRACE_WITH_REGS */
/*
diff --git a/arch/arm64/kernel/ftrace.c b/arch/arm64/kernel/ftrace.c
index 4506c4a90ac1..35eb7c9b5e53 100644
--- a/arch/arm64/kernel/ftrace.c
+++ b/arch/arm64/kernel/ftrace.c
@@ -268,6 +268,22 @@ void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent,
}
#ifdef CONFIG_DYNAMIC_FTRACE
+
+#ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
+void ftrace_graph_func(unsigned long ip, unsigned long parent_ip,
+ struct ftrace_ops *op, struct ftrace_regs *fregs)
+{
+ /*
+ * Athough graph_ops doesn't have FTRACE_OPS_FL_SAVE_REGS set in flags,
+ * regs can't be NULL in DYNAMIC_FTRACE_WITH_REGS. By design, it should
+ * be fixed when DYNAMIC_FTRACE_WITH_ARGS is implemented.
+ */
+ struct pt_regs *regs = arch_ftrace_get_regs(fregs);
+ unsigned long *parent = (unsigned long *)&procedure_link_pointer(regs);
+
+ prepare_ftrace_return(ip, parent, frame_pointer(regs));
+}
+#else
/*
* Turn on/off the call to ftrace_graph_caller() in ftrace_caller()
* depending on @enable.
@@ -297,5 +313,6 @@ int ftrace_disable_ftrace_graph_caller(void)
{
return ftrace_modify_graph_caller(false);
}
+#endif /* CONFIG_DYNAMIC_FTRACE_WITH_REGS */
#endif /* CONFIG_DYNAMIC_FTRACE */
#endif /* CONFIG_FUNCTION_GRAPH_TRACER */
--
2.35.1
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable
2022-04-09 15:35 [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable Chengming Zhou
2022-04-09 15:35 ` [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly Chengming Zhou
@ 2022-04-18 13:24 ` Chengming Zhou
2022-04-19 11:41 ` Mark Rutland
2 siblings, 0 replies; 7+ messages in thread
From: Chengming Zhou @ 2022-04-18 13:24 UTC (permalink / raw)
To: rostedt, mingo, catalin.marinas, will, tglx, bp, dave.hansen,
x86, hpa, broonie, mark.rutland, ardb, linux-arm-kernel
Cc: linux-kernel, duanxiongchun, songmuchun
Hello, friendly ping :-)
On 2022/4/9 23:35, Chengming Zhou wrote:
> The ftrace_[enable,disable]_ftrace_graph_caller() are used to do
> special hooks for graph tracer, which are not needed on some ARCHs
> that use graph_ops:func function to install return_hooker.
>
> So introduce the weak version in ftrace core code to cleanup
> in x86.
>
> Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
> ---
> v4:
> - put weak ftrace_enable,disable_ftrace_graph_caller() in
> fgraph.c instead of ftrace.c as suggested by Steve.
>
> v3:
> - consolidate two #if into a single #if, suggested by Steve. Thanks.
> ---
> arch/x86/kernel/ftrace.c | 17 ++---------------
> kernel/trace/fgraph.c | 18 ++++++++++++++++++
> 2 files changed, 20 insertions(+), 15 deletions(-)
>
> diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
> index 1e31c7d21597..b09d73c2ba89 100644
> --- a/arch/x86/kernel/ftrace.c
> +++ b/arch/x86/kernel/ftrace.c
> @@ -579,9 +579,7 @@ void arch_ftrace_trampoline_free(struct ftrace_ops *ops)
>
> #ifdef CONFIG_FUNCTION_GRAPH_TRACER
>
> -#ifdef CONFIG_DYNAMIC_FTRACE
> -
> -#ifndef CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS
> +#if defined(CONFIG_DYNAMIC_FTRACE) && !defined(CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS)
> extern void ftrace_graph_call(void);
> static const char *ftrace_jmp_replace(unsigned long ip, unsigned long addr)
> {
> @@ -610,18 +608,7 @@ int ftrace_disable_ftrace_graph_caller(void)
>
> return ftrace_mod_jmp(ip, &ftrace_stub);
> }
> -#else /* !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
> -int ftrace_enable_ftrace_graph_caller(void)
> -{
> - return 0;
> -}
> -
> -int ftrace_disable_ftrace_graph_caller(void)
> -{
> - return 0;
> -}
> -#endif /* CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
> -#endif /* !CONFIG_DYNAMIC_FTRACE */
> +#endif /* CONFIG_DYNAMIC_FTRACE && !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
>
> /*
> * Hook the return address and push it in the stack of return addrs
> diff --git a/kernel/trace/fgraph.c b/kernel/trace/fgraph.c
> index 8f4fb328133a..289311680c29 100644
> --- a/kernel/trace/fgraph.c
> +++ b/kernel/trace/fgraph.c
> @@ -30,6 +30,24 @@ int ftrace_graph_active;
> /* Both enabled by default (can be cleared by function_graph tracer flags */
> static bool fgraph_sleep_time = true;
>
> +/*
> + * archs can override this function if they must do something
> + * to enable hook for graph tracer.
> + */
> +int __weak ftrace_enable_ftrace_graph_caller(void)
> +{
> + return 0;
> +}
> +
> +/*
> + * archs can override this function if they must do something
> + * to disable hook for graph tracer.
> + */
> +int __weak ftrace_disable_ftrace_graph_caller(void)
> +{
> + return 0;
> +}
> +
> /**
> * ftrace_graph_stop - set to permanently disable function graph tracing
> *
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable
2022-04-09 15:35 [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable Chengming Zhou
2022-04-09 15:35 ` [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly Chengming Zhou
2022-04-18 13:24 ` [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable Chengming Zhou
@ 2022-04-19 11:41 ` Mark Rutland
2022-04-20 15:30 ` [External] " Chengming Zhou
2 siblings, 1 reply; 7+ messages in thread
From: Mark Rutland @ 2022-04-19 11:41 UTC (permalink / raw)
To: Chengming Zhou
Cc: rostedt, mingo, catalin.marinas, will, tglx, bp, dave.hansen,
x86, hpa, broonie, ardb, linux-arm-kernel, linux-kernel,
duanxiongchun, songmuchun
On Sat, Apr 09, 2022 at 11:35:53PM +0800, Chengming Zhou wrote:
> The ftrace_[enable,disable]_ftrace_graph_caller() are used to do
> special hooks for graph tracer, which are not needed on some ARCHs
> that use graph_ops:func function to install return_hooker.
>
> So introduce the weak version in ftrace core code to cleanup
> in x86.
>
> Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
> ---
> v4:
> - put weak ftrace_enable,disable_ftrace_graph_caller() in
> fgraph.c instead of ftrace.c as suggested by Steve.
>
> v3:
> - consolidate two #if into a single #if, suggested by Steve. Thanks.
> ---
> arch/x86/kernel/ftrace.c | 17 ++---------------
> kernel/trace/fgraph.c | 18 ++++++++++++++++++
> 2 files changed, 20 insertions(+), 15 deletions(-)
>
> diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
> index 1e31c7d21597..b09d73c2ba89 100644
> --- a/arch/x86/kernel/ftrace.c
> +++ b/arch/x86/kernel/ftrace.c
> @@ -579,9 +579,7 @@ void arch_ftrace_trampoline_free(struct ftrace_ops *ops)
>
> #ifdef CONFIG_FUNCTION_GRAPH_TRACER
>
> -#ifdef CONFIG_DYNAMIC_FTRACE
> -
> -#ifndef CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS
> +#if defined(CONFIG_DYNAMIC_FTRACE) && !defined(CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS)
> extern void ftrace_graph_call(void);
> static const char *ftrace_jmp_replace(unsigned long ip, unsigned long addr)
> {
> @@ -610,18 +608,7 @@ int ftrace_disable_ftrace_graph_caller(void)
>
> return ftrace_mod_jmp(ip, &ftrace_stub);
> }
> -#else /* !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
> -int ftrace_enable_ftrace_graph_caller(void)
> -{
> - return 0;
> -}
> -
> -int ftrace_disable_ftrace_graph_caller(void)
> -{
> - return 0;
> -}
> -#endif /* CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
> -#endif /* !CONFIG_DYNAMIC_FTRACE */
> +#endif /* CONFIG_DYNAMIC_FTRACE && !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
>
> /*
> * Hook the return address and push it in the stack of return addrs
> diff --git a/kernel/trace/fgraph.c b/kernel/trace/fgraph.c
> index 8f4fb328133a..289311680c29 100644
> --- a/kernel/trace/fgraph.c
> +++ b/kernel/trace/fgraph.c
> @@ -30,6 +30,24 @@ int ftrace_graph_active;
> /* Both enabled by default (can be cleared by function_graph tracer flags */
> static bool fgraph_sleep_time = true;
>
> +/*
> + * archs can override this function if they must do something
> + * to enable hook for graph tracer.
> + */
> +int __weak ftrace_enable_ftrace_graph_caller(void)
> +{
> + return 0;
> +}
> +
> +/*
> + * archs can override this function if they must do something
> + * to disable hook for graph tracer.
> + */
> +int __weak ftrace_disable_ftrace_graph_caller(void)
> +{
> + return 0;
> +}
IIUC an arch should either:
* Have ftrace_graph_call()
* Have both ftrace_enable_ftrace_graph_caller() and
ftrace_disable_ftrace_graph_caller()
... and I can't think of a reason an arch would need both ftrace_graph_call()
*and* the enable/disable functions.
Given that, could we drop the `__weak` and place these within ifdeffery, i.e.
make the above:
| #ifndef ftrace_graph_call
| int ftrace_enable_ftrace_graph_caller(void) { return 0; }
| int ftrace_disable_ftrace_graph_caller(void) { return 0; }
| #endif /* ftrace_graph_call *.
That way we'd catch cases when:
* An architecture meant to provide one of these functions, but forgot (e.g. the
name got typo'd)
* An architecture provides an unnecessary implementation of either of these
functions.
Regardless, this looks ok to me. Steve, are you happy with this? I suspect we'd
need to take this via the arm64 tree with the next patch, so we'd need your Ack.
Thanks,
Mark.
> +
> /**
> * ftrace_graph_stop - set to permanently disable function graph tracing
> *
> --
> 2.35.1
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly
2022-04-09 15:35 ` [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly Chengming Zhou
@ 2022-04-19 12:55 ` Mark Rutland
2022-04-19 13:27 ` [External] " Chengming Zhou
0 siblings, 1 reply; 7+ messages in thread
From: Mark Rutland @ 2022-04-19 12:55 UTC (permalink / raw)
To: Chengming Zhou
Cc: rostedt, mingo, catalin.marinas, will, tglx, bp, dave.hansen,
x86, hpa, broonie, ardb, linux-arm-kernel, linux-kernel,
duanxiongchun, songmuchun
On Sat, Apr 09, 2022 at 11:35:54PM +0800, Chengming Zhou wrote:
> As we do in commit 0c0593b45c9b ("x86/ftrace: Make function graph
> use ftrace directly"), we don't need special hook for graph tracer,
> but instead we use graph_ops:func function to install return_hooker.
>
> Since commit 3b23e4991fb6 ("arm64: implement ftrace with regs") add
> implementation for FTRACE_WITH_REGS on arm64, we can easily adopt
> the same cleanup on arm64.
>
> And this cleanup only changes the FTRACE_WITH_REGS implementation,
> so the mcount-based implementation is unaffected.
Could you please say *why* we only do this for FTRACE_WITH_REGS? IIUC that's
possible, but would require more invasive refactoring of the core code; have I
understood correctly?
If so, could we please make this:
| While in theory it would be possible to make a similar cleanup for
| !FTRACE_WITH_REGS, this will require rework of the core code, and so for now
| we only change the FTRACE_WITH_REGS implementation.
It'd be quite nice if we could clean up the !FTRACE_WITH_REGS case similarly,
but as it appeass that would require far more invasive changes, I'm happy to
leave that for future work.
>
> Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
> ---
> v3:
> - Add comments in ftrace_graph_func() as suggested by Steve. Thanks.
>
> v2:
> - Remove FTRACE_WITH_REGS ftrace_graph_caller asm, thanks Mark.
> ---
> arch/arm64/include/asm/ftrace.h | 7 +++++++
> arch/arm64/kernel/entry-ftrace.S | 17 -----------------
> arch/arm64/kernel/ftrace.c | 17 +++++++++++++++++
> 3 files changed, 24 insertions(+), 17 deletions(-)
>
> diff --git a/arch/arm64/include/asm/ftrace.h b/arch/arm64/include/asm/ftrace.h
> index 1494cfa8639b..dbc45a4157fa 100644
> --- a/arch/arm64/include/asm/ftrace.h
> +++ b/arch/arm64/include/asm/ftrace.h
> @@ -80,8 +80,15 @@ static inline unsigned long ftrace_call_adjust(unsigned long addr)
>
> #ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
> struct dyn_ftrace;
> +struct ftrace_ops;
> +struct ftrace_regs;
> +
> int ftrace_init_nop(struct module *mod, struct dyn_ftrace *rec);
> #define ftrace_init_nop ftrace_init_nop
> +
> +void ftrace_graph_func(unsigned long ip, unsigned long parent_ip,
> + struct ftrace_ops *op, struct ftrace_regs *fregs);
> +#define ftrace_graph_func ftrace_graph_func
> #endif
>
> #define ftrace_return_address(n) return_address(n)
> diff --git a/arch/arm64/kernel/entry-ftrace.S b/arch/arm64/kernel/entry-ftrace.S
> index e535480a4069..d42a205ef625 100644
> --- a/arch/arm64/kernel/entry-ftrace.S
> +++ b/arch/arm64/kernel/entry-ftrace.S
> @@ -97,12 +97,6 @@ SYM_CODE_START(ftrace_common)
> SYM_INNER_LABEL(ftrace_call, SYM_L_GLOBAL)
> bl ftrace_stub
>
> -#ifdef CONFIG_FUNCTION_GRAPH_TRACER
> -SYM_INNER_LABEL(ftrace_graph_call, SYM_L_GLOBAL) // ftrace_graph_caller();
> - nop // If enabled, this will be replaced
> - // "b ftrace_graph_caller"
> -#endif
> -
> /*
> * At the callsite x0-x8 and x19-x30 were live. Any C code will have preserved
> * x19-x29 per the AAPCS, and we created frame records upon entry, so we need
> @@ -127,17 +121,6 @@ ftrace_common_return:
> ret x9
> SYM_CODE_END(ftrace_common)
>
> -#ifdef CONFIG_FUNCTION_GRAPH_TRACER
> -SYM_CODE_START(ftrace_graph_caller)
> - ldr x0, [sp, #S_PC]
> - sub x0, x0, #AARCH64_INSN_SIZE // ip (callsite's BL insn)
> - add x1, sp, #S_LR // parent_ip (callsite's LR)
> - ldr x2, [sp, #PT_REGS_SIZE] // parent fp (callsite's FP)
> - bl prepare_ftrace_return
> - b ftrace_common_return
> -SYM_CODE_END(ftrace_graph_caller)
> -#endif
> -
> #else /* CONFIG_DYNAMIC_FTRACE_WITH_REGS */
>
> /*
> diff --git a/arch/arm64/kernel/ftrace.c b/arch/arm64/kernel/ftrace.c
> index 4506c4a90ac1..35eb7c9b5e53 100644
> --- a/arch/arm64/kernel/ftrace.c
> +++ b/arch/arm64/kernel/ftrace.c
> @@ -268,6 +268,22 @@ void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent,
> }
>
> #ifdef CONFIG_DYNAMIC_FTRACE
> +
> +#ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
> +void ftrace_graph_func(unsigned long ip, unsigned long parent_ip,
> + struct ftrace_ops *op, struct ftrace_regs *fregs)
> +{
> + /*
> + * Athough graph_ops doesn't have FTRACE_OPS_FL_SAVE_REGS set in flags,
> + * regs can't be NULL in DYNAMIC_FTRACE_WITH_REGS. By design, it should
> + * be fixed when DYNAMIC_FTRACE_WITH_ARGS is implemented.
> + */
This is a bit confusing, since it makes it sound like there's an bug in the
current implementation, rather than something that would need to change if
support for DYNAMIC_FTRACE_WITH_ARGS is added.
Could we please make this:
/*
* When DYNAMIC_FTRACE_WITH_REGS is selected, `fregs` can never be NULL
* and arch_ftrace_get_regs(fregs) will always give a non-NULL pt_regs
* in which we can safely modify the LR.
*/
Other than that, this looks good to me. I gave it a spin under QEMU atop
v5.18-rc3. The CONFIG_FTRACE_STARTUP_TEST tests all pass, and I played with the
graph tracer with:
| # echo do_el0_svc > /sys/kernel/tracing/set_graph_function
| # echo function_graph > /sys/kernel/tracing/current_tracer
... for which the resutls looks sane.
To make sure this didn't adversely affect the return address rewriting, I also
concurrently ran perf with:
| # perf record -g -e raw_syscalls:sys_enter:k /bin/true
| # perf report
... for which the results also looked fine.
I also tested the !DYNAMIC_FTRACE_WITH_REGS modes by building with an older
compiler and also building with !DYNAMIC_FTRACE, which all looked good.
So FWIW:
Tested-by: Mark Rutland <mark.rutland@arm.com>
... and if you make the changes I requested above:
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
If you could spin a v5 with that folded in, that would be great.
Thanks,
Mark.
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [External] Re: [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly
2022-04-19 12:55 ` Mark Rutland
@ 2022-04-19 13:27 ` Chengming Zhou
0 siblings, 0 replies; 7+ messages in thread
From: Chengming Zhou @ 2022-04-19 13:27 UTC (permalink / raw)
To: Mark Rutland
Cc: rostedt, mingo, catalin.marinas, will, tglx, bp, dave.hansen,
x86, hpa, broonie, ardb, linux-arm-kernel, linux-kernel,
duanxiongchun, songmuchun
On 2022/4/19 20:55, Mark Rutland wrote:
> On Sat, Apr 09, 2022 at 11:35:54PM +0800, Chengming Zhou wrote:
>> As we do in commit 0c0593b45c9b ("x86/ftrace: Make function graph
>> use ftrace directly"), we don't need special hook for graph tracer,
>> but instead we use graph_ops:func function to install return_hooker.
>>
>> Since commit 3b23e4991fb6 ("arm64: implement ftrace with regs") add
>> implementation for FTRACE_WITH_REGS on arm64, we can easily adopt
>> the same cleanup on arm64.
>>
>> And this cleanup only changes the FTRACE_WITH_REGS implementation,
>> so the mcount-based implementation is unaffected.
>
> Could you please say *why* we only do this for FTRACE_WITH_REGS? IIUC that's
> possible, but would require more invasive refactoring of the core code; have I
> understood correctly?
Yes, I think so. The static mcount-based implementation should also be changed
in this way, but I haven't look too deep into that asm implementation yet.
>
> If so, could we please make this:
>
> | While in theory it would be possible to make a similar cleanup for
> | !FTRACE_WITH_REGS, this will require rework of the core code, and so for now
> | we only change the FTRACE_WITH_REGS implementation.
>
> It'd be quite nice if we could clean up the !FTRACE_WITH_REGS case similarly,
> but as it appeass that would require far more invasive changes, I'm happy to
> leave that for future work.
Ok, will add it in the commit message. And leave this for future work.
>
>>
>> Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
>> ---
>> v3:
>> - Add comments in ftrace_graph_func() as suggested by Steve. Thanks.
>>
>> v2:
>> - Remove FTRACE_WITH_REGS ftrace_graph_caller asm, thanks Mark.
>> ---
>> arch/arm64/include/asm/ftrace.h | 7 +++++++
>> arch/arm64/kernel/entry-ftrace.S | 17 -----------------
>> arch/arm64/kernel/ftrace.c | 17 +++++++++++++++++
>> 3 files changed, 24 insertions(+), 17 deletions(-)
>>
>> diff --git a/arch/arm64/include/asm/ftrace.h b/arch/arm64/include/asm/ftrace.h
>> index 1494cfa8639b..dbc45a4157fa 100644
>> --- a/arch/arm64/include/asm/ftrace.h
>> +++ b/arch/arm64/include/asm/ftrace.h
>> @@ -80,8 +80,15 @@ static inline unsigned long ftrace_call_adjust(unsigned long addr)
>>
>> #ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
>> struct dyn_ftrace;
>> +struct ftrace_ops;
>> +struct ftrace_regs;
>> +
>> int ftrace_init_nop(struct module *mod, struct dyn_ftrace *rec);
>> #define ftrace_init_nop ftrace_init_nop
>> +
>> +void ftrace_graph_func(unsigned long ip, unsigned long parent_ip,
>> + struct ftrace_ops *op, struct ftrace_regs *fregs);
>> +#define ftrace_graph_func ftrace_graph_func
>> #endif
>>
>> #define ftrace_return_address(n) return_address(n)
>> diff --git a/arch/arm64/kernel/entry-ftrace.S b/arch/arm64/kernel/entry-ftrace.S
>> index e535480a4069..d42a205ef625 100644
>> --- a/arch/arm64/kernel/entry-ftrace.S
>> +++ b/arch/arm64/kernel/entry-ftrace.S
>> @@ -97,12 +97,6 @@ SYM_CODE_START(ftrace_common)
>> SYM_INNER_LABEL(ftrace_call, SYM_L_GLOBAL)
>> bl ftrace_stub
>>
>> -#ifdef CONFIG_FUNCTION_GRAPH_TRACER
>> -SYM_INNER_LABEL(ftrace_graph_call, SYM_L_GLOBAL) // ftrace_graph_caller();
>> - nop // If enabled, this will be replaced
>> - // "b ftrace_graph_caller"
>> -#endif
>> -
>> /*
>> * At the callsite x0-x8 and x19-x30 were live. Any C code will have preserved
>> * x19-x29 per the AAPCS, and we created frame records upon entry, so we need
>> @@ -127,17 +121,6 @@ ftrace_common_return:
>> ret x9
>> SYM_CODE_END(ftrace_common)
>>
>> -#ifdef CONFIG_FUNCTION_GRAPH_TRACER
>> -SYM_CODE_START(ftrace_graph_caller)
>> - ldr x0, [sp, #S_PC]
>> - sub x0, x0, #AARCH64_INSN_SIZE // ip (callsite's BL insn)
>> - add x1, sp, #S_LR // parent_ip (callsite's LR)
>> - ldr x2, [sp, #PT_REGS_SIZE] // parent fp (callsite's FP)
>> - bl prepare_ftrace_return
>> - b ftrace_common_return
>> -SYM_CODE_END(ftrace_graph_caller)
>> -#endif
>> -
>> #else /* CONFIG_DYNAMIC_FTRACE_WITH_REGS */
>>
>> /*
>> diff --git a/arch/arm64/kernel/ftrace.c b/arch/arm64/kernel/ftrace.c
>> index 4506c4a90ac1..35eb7c9b5e53 100644
>> --- a/arch/arm64/kernel/ftrace.c
>> +++ b/arch/arm64/kernel/ftrace.c
>> @@ -268,6 +268,22 @@ void prepare_ftrace_return(unsigned long self_addr, unsigned long *parent,
>> }
>>
>> #ifdef CONFIG_DYNAMIC_FTRACE
>> +
>> +#ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
>> +void ftrace_graph_func(unsigned long ip, unsigned long parent_ip,
>> + struct ftrace_ops *op, struct ftrace_regs *fregs)
>> +{
>> + /*
>> + * Athough graph_ops doesn't have FTRACE_OPS_FL_SAVE_REGS set in flags,
>> + * regs can't be NULL in DYNAMIC_FTRACE_WITH_REGS. By design, it should
>> + * be fixed when DYNAMIC_FTRACE_WITH_ARGS is implemented.
>> + */
>
> This is a bit confusing, since it makes it sound like there's an bug in the
> current implementation, rather than something that would need to change if
> support for DYNAMIC_FTRACE_WITH_ARGS is added.
>
> Could we please make this:
>
> /*
> * When DYNAMIC_FTRACE_WITH_REGS is selected, `fregs` can never be NULL
> * and arch_ftrace_get_regs(fregs) will always give a non-NULL pt_regs
> * in which we can safely modify the LR.
> */
>
Ok, will do. This expression is nicer, the previous comment maybe make people
think it's an bug to be fixed.
> Other than that, this looks good to me. I gave it a spin under QEMU atop
> v5.18-rc3. The CONFIG_FTRACE_STARTUP_TEST tests all pass, and I played with the
> graph tracer with:
>
> | # echo do_el0_svc > /sys/kernel/tracing/set_graph_function
> | # echo function_graph > /sys/kernel/tracing/current_tracer
>
> ... for which the resutls looks sane.
>
> To make sure this didn't adversely affect the return address rewriting, I also
> concurrently ran perf with:
>
> | # perf record -g -e raw_syscalls:sys_enter:k /bin/true
> | # perf report
>
> ... for which the results also looked fine.
>
> I also tested the !DYNAMIC_FTRACE_WITH_REGS modes by building with an older
> compiler and also building with !DYNAMIC_FTRACE, which all looked good.
>
> So FWIW:
>
> Tested-by: Mark Rutland <mark.rutland@arm.com>
>
> ... and if you make the changes I requested above:
>
> Reviewed-by: Mark Rutland <mark.rutland@arm.com>
>
> If you could spin a v5 with that folded in, that would be great.
Of course, will do in v5.
Thanks.
>
> Thanks,
> Mark.
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [External] Re: [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable
2022-04-19 11:41 ` Mark Rutland
@ 2022-04-20 15:30 ` Chengming Zhou
0 siblings, 0 replies; 7+ messages in thread
From: Chengming Zhou @ 2022-04-20 15:30 UTC (permalink / raw)
To: Mark Rutland, rostedt
Cc: mingo, catalin.marinas, will, tglx, bp, dave.hansen, x86, hpa,
broonie, ardb, linux-arm-kernel, linux-kernel, duanxiongchun,
songmuchun
Hi,
On 2022/4/19 19:41, Mark Rutland wrote:
> On Sat, Apr 09, 2022 at 11:35:53PM +0800, Chengming Zhou wrote:
>> The ftrace_[enable,disable]_ftrace_graph_caller() are used to do
>> special hooks for graph tracer, which are not needed on some ARCHs
>> that use graph_ops:func function to install return_hooker.
>>
>> So introduce the weak version in ftrace core code to cleanup
>> in x86.
>>
>> Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
>> ---
>> v4:
>> - put weak ftrace_enable,disable_ftrace_graph_caller() in
>> fgraph.c instead of ftrace.c as suggested by Steve.
>>
>> v3:
>> - consolidate two #if into a single #if, suggested by Steve. Thanks.
>> ---
>> arch/x86/kernel/ftrace.c | 17 ++---------------
>> kernel/trace/fgraph.c | 18 ++++++++++++++++++
>> 2 files changed, 20 insertions(+), 15 deletions(-)
>>
>> diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
>> index 1e31c7d21597..b09d73c2ba89 100644
>> --- a/arch/x86/kernel/ftrace.c
>> +++ b/arch/x86/kernel/ftrace.c
>> @@ -579,9 +579,7 @@ void arch_ftrace_trampoline_free(struct ftrace_ops *ops)
>>
>> #ifdef CONFIG_FUNCTION_GRAPH_TRACER
>>
>> -#ifdef CONFIG_DYNAMIC_FTRACE
>> -
>> -#ifndef CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS
>> +#if defined(CONFIG_DYNAMIC_FTRACE) && !defined(CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS)
>> extern void ftrace_graph_call(void);
>> static const char *ftrace_jmp_replace(unsigned long ip, unsigned long addr)
>> {
>> @@ -610,18 +608,7 @@ int ftrace_disable_ftrace_graph_caller(void)
>>
>> return ftrace_mod_jmp(ip, &ftrace_stub);
>> }
>> -#else /* !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
>> -int ftrace_enable_ftrace_graph_caller(void)
>> -{
>> - return 0;
>> -}
>> -
>> -int ftrace_disable_ftrace_graph_caller(void)
>> -{
>> - return 0;
>> -}
>> -#endif /* CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
>> -#endif /* !CONFIG_DYNAMIC_FTRACE */
>> +#endif /* CONFIG_DYNAMIC_FTRACE && !CONFIG_HAVE_DYNAMIC_FTRACE_WITH_ARGS */
>>
>> /*
>> * Hook the return address and push it in the stack of return addrs
>> diff --git a/kernel/trace/fgraph.c b/kernel/trace/fgraph.c
>> index 8f4fb328133a..289311680c29 100644
>> --- a/kernel/trace/fgraph.c
>> +++ b/kernel/trace/fgraph.c
>> @@ -30,6 +30,24 @@ int ftrace_graph_active;
>> /* Both enabled by default (can be cleared by function_graph tracer flags */
>> static bool fgraph_sleep_time = true;
>>
>> +/*
>> + * archs can override this function if they must do something
>> + * to enable hook for graph tracer.
>> + */
>> +int __weak ftrace_enable_ftrace_graph_caller(void)
>> +{
>> + return 0;
>> +}
>> +
>> +/*
>> + * archs can override this function if they must do something
>> + * to disable hook for graph tracer.
>> + */
>> +int __weak ftrace_disable_ftrace_graph_caller(void)
>> +{
>> + return 0;
>> +}
>
> IIUC an arch should either:
>
> * Have ftrace_graph_call()
>
> * Have both ftrace_enable_ftrace_graph_caller() and
> ftrace_disable_ftrace_graph_caller()
>
> ... and I can't think of a reason an arch would need both ftrace_graph_call()
> *and* the enable/disable functions.
This way is more precise, but we have to add "#define ftrace_graph_call ftrace_graph_call"
in all ARCH's asm/ftrace.h correctly.
Previously we only need to define override ftrace_[enable,disable]_ftrace_graph_caller()
in ARCH's ftrace.c, looks like a little more complexity?
>
> Given that, could we drop the `__weak` and place these within ifdeffery, i.e.
> make the above:
>
> | #ifndef ftrace_graph_call
> | int ftrace_enable_ftrace_graph_caller(void) { return 0; }
> | int ftrace_disable_ftrace_graph_caller(void) { return 0; }
> | #endif /* ftrace_graph_call *.
>
> That way we'd catch cases when:
>
> * An architecture meant to provide one of these functions, but forgot (e.g. the
> name got typo'd)
>
> * An architecture provides an unnecessary implementation of either of these
> functions.
I tried this way today, and it works too. I'm ok with both way. Maybe I should
send v5 with this patch unchanged first. Looking forward to Steve's opinion.
Thanks.
>
> Regardless, this looks ok to me. Steve, are you happy with this? I suspect we'd
> need to take this via the arm64 tree with the next patch, so we'd need your Ack.
>
> Thanks,
> Mark.
>
>> +
>> /**
>> * ftrace_graph_stop - set to permanently disable function graph tracing
>> *
>> --
>> 2.35.1
>>
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2022-04-20 15:30 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-04-09 15:35 [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable Chengming Zhou
2022-04-09 15:35 ` [PATCH v4 2/2] arm64/ftrace: Make function graph use ftrace directly Chengming Zhou
2022-04-19 12:55 ` Mark Rutland
2022-04-19 13:27 ` [External] " Chengming Zhou
2022-04-18 13:24 ` [PATCH v4 1/2] ftrace: cleanup ftrace_graph_caller enable and disable Chengming Zhou
2022-04-19 11:41 ` Mark Rutland
2022-04-20 15:30 ` [External] " Chengming Zhou
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).