Fix the output of ftrace in hex mode. Without this patch, the ouput of ftrace is :
raw mode : 6474 0 141531612444 0 140 + 6402 120 S
hex mode : 000091a4 00000000 000000023f1f50c1 00000000 c8 000000b2 00009120 87 ffff00c8 00000035
There is an inversion on ouput hex(6474) is 194a
Signed-off-by: Philippe Reynes <[email protected]>
---
Index: linux-2.6.26/kernel/trace/trace.c
===================================================================
--- linux-2.6.26/kernel/trace/trace.c
+++ linux-2.6.26/kernel/trace/trace.c 2008-10-08 22:30:29.000000000 +0200
@@ -415,8 +415,8 @@
#endif
byte = data[i];
- hex[j++] = hex2asc[byte & 0x0f];
hex[j++] = hex2asc[byte >> 4];
+ hex[j++] = hex2asc[byte & 0x0f];
}
hex[j++] = ' ';
On Thu, 9 Oct 2008, trem wrote:
> Index: linux-2.6.26/kernel/trace/trace.c
> ===================================================================
> --- linux-2.6.26/kernel/trace/trace.c
> +++ linux-2.6.26/kernel/trace/trace.c 2008-10-08 22:30:29.000000000 +0200
> @@ -415,8 +415,8 @@
> #endif
> byte = data[i];
>
> - hex[j++] = hex2asc[byte & 0x0f];
> hex[j++] = hex2asc[byte >> 4];
> + hex[j++] = hex2asc[byte & 0x0f];
> }
> hex[j++] = ' ';
I'm surprised Harvey Harrison hasn't changed it to pack_hex_byte
and removed the static.
On Wed, 2008-10-08 at 15:22 -0700, Joe Perches wrote:
> I'm surprised Harvey Harrison hasn't changed it to pack_hex_byte
> and removed the static.
>
From: Harvey Harrison <[email protected]>
Subject: [PATCH] ftrace: Fix inversion of hex output and use common routines
Fix the output of ftrace in hex mode as the hi/lo nibbles are output in
reverse order. Without this patch, the output of ftrace is:
raw mode : 6474 0 141531612444 0 140 + 6402 120 S
hex mode : 000091a4 00000000 000000023f1f50c1 00000000 c8 000000b2 00009120 87 ffff00c8 00000035
There is an inversion on ouput hex(6474) is 194a
[based on a patch by Philippe Reynes <[email protected]>]
Signed-off-by: Harvey Harrison <[email protected]>
---
kernel/trace/trace.c | 8 ++------
1 files changed, 2 insertions(+), 6 deletions(-)
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 8f3fb3d..763f763 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -396,14 +396,12 @@ trace_seq_putmem(struct trace_seq *s, void *mem, size_t len)
}
#define HEX_CHARS 17
-static const char hex2asc[] = "0123456789abcdef";
static int
trace_seq_putmem_hex(struct trace_seq *s, void *mem, size_t len)
{
unsigned char hex[HEX_CHARS];
unsigned char *data = mem;
- unsigned char byte;
int i, j;
BUG_ON(len >= HEX_CHARS);
@@ -413,10 +411,8 @@ trace_seq_putmem_hex(struct trace_seq *s, void *mem, size_t len)
#else
for (i = len-1, j = 0; i >= 0; i--) {
#endif
- byte = data[i];
-
- hex[j++] = hex2asc[byte & 0x0f];
- hex[j++] = hex2asc[byte >> 4];
+ hex[j++] = hex_asc_hi(data[i]);
+ hex[j++] = hex_asc_lo(data[i]);
}
hex[j++] = ' ';
--
1.6.0.2.471.g47a76
On Wed, 2008-10-08 at 17:16 -0700, Joe Perches wrote:
> On Wed, 8 Oct 2008, Harvey Harrison wrote:
> > From: Harvey Harrison <[email protected]>
> > Subject: [PATCH] ftrace: Fix inversion of hex output and use common routines
>
> You're quick...
>
> > BUG_ON(len >= HEX_CHARS);
>
> I think the BUG_ON is senseless.
>
Agreed, it probably meant to say BUG_ON(len * 2 > HEX_CHARS -1)
But as this is only called from within a helper macro, it could be
changed to a build-time check instead of a runtime:
From: Harvey Harrison <[email protected]>
Subject: [PATCH] trace: add build-time check to avoid overrunning hex buffer
Remove the runtime BUG_ON and change to a compile-time check in
the macro that calls the hex format routine
[Noticed by Joe Perches]
Signed-off-by: Harvey Harrison <[email protected]>
---
kernel/trace/trace.c | 6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 763f763..1ffbc24 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -395,7 +395,8 @@ trace_seq_putmem(struct trace_seq *s, void *mem, size_t len)
return len;
}
-#define HEX_CHARS 17
+#define MAX_MEMHEX_BYTES (8)
+#define HEX_CHARS ((MAX_MEMHEX_BYTES * 2) + 1)
static int
trace_seq_putmem_hex(struct trace_seq *s, void *mem, size_t len)
@@ -404,8 +405,6 @@ trace_seq_putmem_hex(struct trace_seq *s, void *mem, size_t len)
unsigned char *data = mem;
int i, j;
- BUG_ON(len >= HEX_CHARS);
-
#ifdef __BIG_ENDIAN
for (i = 0, j = 0; i < len; i++) {
#else
@@ -1706,6 +1705,7 @@ do { \
#define SEQ_PUT_HEX_FIELD_RET(s, x) \
do { \
+ BUILD_BUG_ON(sizeof(x) > MAX_MEMHEX_BYTES); \
if (!trace_seq_putmem_hex(s, &(x), sizeof(x))) \
return 0; \
} while (0)
--
1.6.0.2.471.g47a76
* Harvey Harrison <[email protected]> wrote:
> On Wed, 2008-10-08 at 17:16 -0700, Joe Perches wrote:
> > On Wed, 8 Oct 2008, Harvey Harrison wrote:
> > > From: Harvey Harrison <[email protected]>
> > > Subject: [PATCH] ftrace: Fix inversion of hex output and use common routines
> >
> > You're quick...
> >
> > > BUG_ON(len >= HEX_CHARS);
> >
> > I think the BUG_ON is senseless.
> >
>
> Agreed, it probably meant to say BUG_ON(len * 2 > HEX_CHARS -1)
>
> But as this is only called from within a helper macro, it could be
> changed to a build-time check instead of a runtime:
>
> From: Harvey Harrison <[email protected]>
> Subject: [PATCH] trace: add build-time check to avoid overrunning hex buffer
applied to tip/tracing/core:
bb86ba7: trace: add build-time check to avoid overrunning hex buffer
32ce86f: ftrace: fix hex output mode of ftrace
thanks guys!
Ingo
On Thu, 2008-10-09 at 14:31 +0200, Ingo Molnar wrote:
> * Harvey Harrison <[email protected]> wrote:
>
> > On Wed, 2008-10-08 at 17:16 -0700, Joe Perches wrote:
> > > On Wed, 8 Oct 2008, Harvey Harrison wrote:
> > > > From: Harvey Harrison <[email protected]>
> > > > Subject: [PATCH] ftrace: Fix inversion of hex output and use common routines
> > >
> > > You're quick...
> > >
> > > > BUG_ON(len >= HEX_CHARS);
> > >
> > > I think the BUG_ON is senseless.
> > >
> >
> > Agreed, it probably meant to say BUG_ON(len * 2 > HEX_CHARS -1)
> >
> > But as this is only called from within a helper macro, it could be
> > changed to a build-time check instead of a runtime:
> >
> > From: Harvey Harrison <[email protected]>
> > Subject: [PATCH] trace: add build-time check to avoid overrunning hex buffer
>
> applied to tip/tracing/core:
>
> bb86ba7: trace: add build-time check to avoid overrunning hex buffer
> 32ce86f: ftrace: fix hex output mode of ftrace
>
Thinking about this, these should probably be forwarded to -stable once
they go into mainline.
Cheers,
Harvey