From: "Steven Rostedt (Google)" <[email protected]>
Add a test that writes longs strings, some over the size of the sub buffer
and make sure that the entire content is there.
Signed-off-by: Steven Rostedt (Google) <[email protected]>
---
Changes since v2: https://lore.kernel.org/linux-trace-kernel/[email protected]
- Realized with the upcoming change of the dynamic subbuffer sizes, that
this test will fail if the subbuffer is bigger than what the trace_seq
can hold. Now the trace_marker does not always utilize the full subbuffer
but the size of the trace_seq instead. As that size isn't available to
user space, we can only just make sure all content is there.
.../ftrace/test.d/00basic/trace_marker.tc | 82 +++++++++++++++++++
1 file changed, 82 insertions(+)
create mode 100755 tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
diff --git a/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
new file mode 100755
index 000000000000..b24aff5807df
--- /dev/null
+++ b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
@@ -0,0 +1,82 @@
+#!/bin/sh
+# SPDX-License-Identifier: GPL-2.0
+# description: Basic tests on writing to trace_marker
+# requires: trace_marker
+# flags: instance
+
+get_buffer_data_size() {
+ sed -ne 's/^.*data.*size:\([0-9][0-9]*\).*/\1/p' events/header_page
+}
+
+get_buffer_data_offset() {
+ sed -ne 's/^.*data.*offset:\([0-9][0-9]*\).*/\1/p' events/header_page
+}
+
+get_event_header_size() {
+ type_len=`sed -ne 's/^.*type_len.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
+ time_len=`sed -ne 's/^.*time_delta.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
+ array_len=`sed -ne 's/^.*array.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
+ total_bits=$((type_len+time_len+array_len))
+ total_bits=$((total_bits+7))
+ echo $((total_bits/8))
+}
+
+get_print_event_buf_offset() {
+ sed -ne 's/^.*buf.*offset:\([0-9][0-9]*\).*/\1/p' events/ftrace/print/format
+}
+
+event_header_size=`get_event_header_size`
+print_header_size=`get_print_event_buf_offset`
+
+data_offset=`get_buffer_data_offset`
+
+marker_meta=$((event_header_size+print_header_size))
+
+make_str() {
+ cnt=$1
+ # subtract two for \n\0 as marker adds these
+ cnt=$((cnt-2))
+ printf -- 'X%.0s' $(seq $cnt)
+}
+
+write_buffer() {
+ size=$1
+
+ str=`make_str $size`
+
+ # clear the buffer
+ echo > trace
+
+ # write the string into the marker
+ echo -n $str > trace_marker
+
+ echo $str
+}
+
+test_buffer() {
+
+ size=`get_buffer_data_size`
+ oneline_size=$((size-marker_meta))
+ echo size = $size
+ echo meta size = $marker_meta
+
+ # Now add a little more the meta data overhead will overflow
+
+ str=`write_buffer $size`
+
+ # Make sure the line was broken
+ new_str=`awk ' /tracing_mark_write:/ { sub(/^.*tracing_mark_write: /,"");printf "%s", $0; exit}' trace`
+
+ if [ "$new_str" = "$str" ]; then
+ exit fail;
+ fi
+
+ # Make sure the entire line can be found
+ new_str=`awk ' /tracing_mark_write:/ { sub(/^.*tracing_mark_write: */,"");printf "%s", $0; }' trace`
+
+ if [ "$new_str" != "$str" ]; then
+ exit fail;
+ fi
+}
+
+test_buffer
--
2.42.0
The REs used in the sed commands below may be simplified as shown if desired.
On Wed, Dec 13, 2023 at 2:22 AM Steven Rostedt <[email protected]> wrote:
>
> From: "Steven Rostedt (Google)" <[email protected]>
>
> Add a test that writes longs strings, some over the size of the sub buffer
> and make sure that the entire content is there.
>
> Signed-off-by: Steven Rostedt (Google) <[email protected]>
> ---
> Changes since v2: https://lore.kernel.org/linux-trace-kernel/[email protected]
>
> - Realized with the upcoming change of the dynamic subbuffer sizes, that
> this test will fail if the subbuffer is bigger than what the trace_seq
> can hold. Now the trace_marker does not always utilize the full subbuffer
> but the size of the trace_seq instead. As that size isn't available to
> user space, we can only just make sure all content is there.
>
> .../ftrace/test.d/00basic/trace_marker.tc | 82 +++++++++++++++++++
> 1 file changed, 82 insertions(+)
> create mode 100755 tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
>
> diff --git a/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> new file mode 100755
> index 000000000000..b24aff5807df
> --- /dev/null
> +++ b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> @@ -0,0 +1,82 @@
> +#!/bin/sh
> +# SPDX-License-Identifier: GPL-2.0
> +# description: Basic tests on writing to trace_marker
> +# requires: trace_marker
> +# flags: instance
> +
> +get_buffer_data_size() {
> + sed -ne 's/^.*data.*size:\([0-9][0-9]*\).*/\1/p' events/header_page
sed -n 's!.*data.*size:\([^;]*\).*!\1!p' events/header_page
> +}
> +
> +get_buffer_data_offset() {
> + sed -ne 's/^.*data.*offset:\([0-9][0-9]*\).*/\1/p' events/header_page
sed -n 's!.*data.*offset:\([^;]*\).*!\1!p' events/header_page
> +}
> +
> +get_event_header_size() {
> + type_len=`sed -ne 's/^.*type_len.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
type_len=`sed -n '/type_len.*bits/s![^0-9]*!!gp'
events/header_event`
> + time_len=`sed -ne 's/^.*time_delta.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
time_len=`sed -n '/time_delta/s![^0-9]*!!gp' events/header_event`
> + array_len=`sed -ne 's/^.*array.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
array_len=`sed -n '/array/s![^0-9]*!!gp' events/header_event`
> + total_bits=$((type_len+time_len+array_len))
> + total_bits=$((total_bits+7))
> + echo $((total_bits/8))
> +}
> +
> +get_print_event_buf_offset() {
> + sed -ne 's/^.*buf.*offset:\([0-9][0-9]*\).*/\1/p' events/ftrace/print/format
sed -n 's!.*buf.*offset:\([^;]*\).*!\1!p' events/ftrace/print/format
> +}
> +
> +event_header_size=`get_event_header_size`
> +print_header_size=`get_print_event_buf_offset`
> +
> +data_offset=`get_buffer_data_offset`
> +
> +marker_meta=$((event_header_size+print_header_size))
> +
> +make_str() {
> + cnt=$1
> + # subtract two for \n\0 as marker adds these
> + cnt=$((cnt-2))
> + printf -- 'X%.0s' $(seq $cnt)
> +}
> +
> +write_buffer() {
> + size=$1
> +
> + str=`make_str $size`
> +
> + # clear the buffer
> + echo > trace
> +
> + # write the string into the marker
> + echo -n $str > trace_marker
> +
> + echo $str
> +}
> +
> +test_buffer() {
> +
> + size=`get_buffer_data_size`
> + oneline_size=$((size-marker_meta))
> + echo size = $size
> + echo meta size = $marker_meta
> +
> + # Now add a little more the meta data overhead will overflow
> +
> + str=`write_buffer $size`
> +
> + # Make sure the line was broken
> + new_str=`awk ' /tracing_mark_write:/ { sub(/^.*tracing_mark_write: /,"");printf "%s", $0; exit}' trace`
> +
> + if [ "$new_str" = "$str" ]; then
> + exit fail;
> + fi
> +
> + # Make sure the entire line can be found
> + new_str=`awk ' /tracing_mark_write:/ { sub(/^.*tracing_mark_write: */,"");printf "%s", $0; }' trace`
> +
> + if [ "$new_str" != "$str" ]; then
> + exit fail;
> + fi
> +}
> +
> +test_buffer
> --
> 2.42.0
>
On Wed, 13 Dec 2023 10:09:50 +0200
Alexander Kapshuk <[email protected]> wrote:
> The REs used in the sed commands below may be simplified as shown if desired.
>
> On Wed, Dec 13, 2023 at 2:22 AM Steven Rostedt <[email protected]> wrote:
> >
> > From: "Steven Rostedt (Google)" <[email protected]>
> >
> > Add a test that writes longs strings, some over the size of the sub buffer
> > and make sure that the entire content is there.
> >
> > Signed-off-by: Steven Rostedt (Google) <[email protected]>
> > ---
> > Changes since v2: https://lore.kernel.org/linux-trace-kernel/[email protected]
> >
> > - Realized with the upcoming change of the dynamic subbuffer sizes, that
> > this test will fail if the subbuffer is bigger than what the trace_seq
> > can hold. Now the trace_marker does not always utilize the full subbuffer
> > but the size of the trace_seq instead. As that size isn't available to
> > user space, we can only just make sure all content is there.
> >
> > .../ftrace/test.d/00basic/trace_marker.tc | 82 +++++++++++++++++++
> > 1 file changed, 82 insertions(+)
> > create mode 100755 tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> >
> > diff --git a/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> > new file mode 100755
> > index 000000000000..b24aff5807df
> > --- /dev/null
> > +++ b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> > @@ -0,0 +1,82 @@
> > +#!/bin/sh
> > +# SPDX-License-Identifier: GPL-2.0
> > +# description: Basic tests on writing to trace_marker
> > +# requires: trace_marker
> > +# flags: instance
> > +
> > +get_buffer_data_size() {
> > + sed -ne 's/^.*data.*size:\([0-9][0-9]*\).*/\1/p' events/header_page
> sed -n 's!.*data.*size:\([^;]*\).*!\1!p' events/header_page
Not sure the above change can be considered simpler, but it also loses out
showing what exactly is being done.
With the original, I have:
sed -ne 's/^.*data.*size:\([0-9][0-9]*\).*/\1/p' events/header_page
Which is obvious that I'm grabbing a number for the size field.
sed -n 's!.*data.*size:\([^;]*\).*!\1!p' events/header_page
Shows that I'm grabbing something after size.
> > +}
> > +
> > +get_buffer_data_offset() {
> > + sed -ne 's/^.*data.*offset:\([0-9][0-9]*\).*/\1/p' events/header_page
> sed -n 's!.*data.*offset:\([^;]*\).*!\1!p' events/header_page
Same here.
> > +}
> > +
> > +get_event_header_size() {
> > + type_len=`sed -ne 's/^.*type_len.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
> type_len=`sed -n '/type_len.*bits/s![^0-9]*!!gp'
> events/header_event`
Honestly, the above may be "simplier" but I can't make out what exactly
that line is doing without going back and looking at the text that's in the
format field.
>
> > + time_len=`sed -ne 's/^.*time_delta.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
> time_len=`sed -n '/time_delta/s![^0-9]*!!gp' events/header_event`
>
> > + array_len=`sed -ne 's/^.*array.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
> array_len=`sed -n '/array/s![^0-9]*!!gp' events/header_event`
>
> > + total_bits=$((type_len+time_len+array_len))
> > + total_bits=$((total_bits+7))
> > + echo $((total_bits/8))
> > +}
> > +
> > +get_print_event_buf_offset() {
> > + sed -ne 's/^.*buf.*offset:\([0-9][0-9]*\).*/\1/p' events/ftrace/print/format
> sed -n 's!.*buf.*offset:\([^;]*\).*!\1!p' events/ftrace/print/format
> > +}
> > +
Yeah, thanks for the suggestions, but I rather have it be more readable
than "simplified". I write perl code the same way. I do not write it like
any perl developer would, because I write it like C code. I want my code to
be easily understandable.
RE can become extremely obfuscated. So, even though my REs are not the
simplest, they tend to be rather easy to understand what they are doing,
and why.
Cheers,
-- Steve
On Wed, Dec 13, 2023 at 4:57 PM Steven Rostedt <[email protected]> wrote:
>
> On Wed, 13 Dec 2023 10:09:50 +0200
> Alexander Kapshuk <[email protected]> wrote:
>
> > The REs used in the sed commands below may be simplified as shown if desired.
> >
> > On Wed, Dec 13, 2023 at 2:22 AM Steven Rostedt <[email protected]> wrote:
> > >
> > > From: "Steven Rostedt (Google)" <[email protected]>
> > >
> > > Add a test that writes longs strings, some over the size of the sub buffer
> > > and make sure that the entire content is there.
> > >
> > > Signed-off-by: Steven Rostedt (Google) <[email protected]>
> > > ---
> > > Changes since v2: https://lore.kernel.org/linux-trace-kernel/[email protected]
> > >
> > > - Realized with the upcoming change of the dynamic subbuffer sizes, that
> > > this test will fail if the subbuffer is bigger than what the trace_seq
> > > can hold. Now the trace_marker does not always utilize the full subbuffer
> > > but the size of the trace_seq instead. As that size isn't available to
> > > user space, we can only just make sure all content is there.
> > >
> > > .../ftrace/test.d/00basic/trace_marker.tc | 82 +++++++++++++++++++
> > > 1 file changed, 82 insertions(+)
> > > create mode 100755 tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> > >
> > > diff --git a/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> > > new file mode 100755
> > > index 000000000000..b24aff5807df
> > > --- /dev/null
> > > +++ b/tools/testing/selftests/ftrace/test.d/00basic/trace_marker.tc
> > > @@ -0,0 +1,82 @@
> > > +#!/bin/sh
> > > +# SPDX-License-Identifier: GPL-2.0
> > > +# description: Basic tests on writing to trace_marker
> > > +# requires: trace_marker
> > > +# flags: instance
> > > +
> > > +get_buffer_data_size() {
> > > + sed -ne 's/^.*data.*size:\([0-9][0-9]*\).*/\1/p' events/header_page
> > sed -n 's!.*data.*size:\([^;]*\).*!\1!p' events/header_page
>
> Not sure the above change can be considered simpler, but it also loses out
> showing what exactly is being done.
>
> With the original, I have:
>
> sed -ne 's/^.*data.*size:\([0-9][0-9]*\).*/\1/p' events/header_page
>
> Which is obvious that I'm grabbing a number for the size field.
>
> sed -n 's!.*data.*size:\([^;]*\).*!\1!p' events/header_page
>
> Shows that I'm grabbing something after size.
>
>
>
> > > +}
> > > +
> > > +get_buffer_data_offset() {
> > > + sed -ne 's/^.*data.*offset:\([0-9][0-9]*\).*/\1/p' events/header_page
> > sed -n 's!.*data.*offset:\([^;]*\).*!\1!p' events/header_page
>
> Same here.
>
> > > +}
> > > +
> > > +get_event_header_size() {
> > > + type_len=`sed -ne 's/^.*type_len.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
> > type_len=`sed -n '/type_len.*bits/s![^0-9]*!!gp'
> > events/header_event`
>
> Honestly, the above may be "simplier" but I can't make out what exactly
> that line is doing without going back and looking at the text that's in the
> format field.
>
> >
> > > + time_len=`sed -ne 's/^.*time_delta.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
> > time_len=`sed -n '/time_delta/s![^0-9]*!!gp' events/header_event`
> >
> > > + array_len=`sed -ne 's/^.*array.*:[^0-9]*\([0-9][0-9]*\).*/\1/p' events/header_event`
> > array_len=`sed -n '/array/s![^0-9]*!!gp' events/header_event`
> >
> > > + total_bits=$((type_len+time_len+array_len))
> > > + total_bits=$((total_bits+7))
> > > + echo $((total_bits/8))
> > > +}
> > > +
> > > +get_print_event_buf_offset() {
> > > + sed -ne 's/^.*buf.*offset:\([0-9][0-9]*\).*/\1/p' events/ftrace/print/format
> > sed -n 's!.*buf.*offset:\([^;]*\).*!\1!p' events/ftrace/print/format
> > > +}
> > > +
>
> Yeah, thanks for the suggestions, but I rather have it be more readable
> than "simplified". I write perl code the same way. I do not write it like
> any perl developer would, because I write it like C code. I want my code to
> be easily understandable.
>
> RE can become extremely obfuscated. So, even though my REs are not the
> simplest, they tend to be rather easy to understand what they are doing,
> and why.
>
> Cheers,
>
> -- Steve
Fair enough.
Thanks.