Test cases which use echo to write strings containing backslashes fail with some shells, as echo's treatment of backslashes in strings varies between shell implementations. Use printf instead, as it should behave consistently across different shells. This requires adjustments to the strings to escape \ and % characters. ftrace_errlog_check() must also re-escape these characters after processing them to remove ^ characters.
Signed-off-by: Seth Forshee seth.forshee@canonical.com --- Changes in v2: - Escape backslashes for a couple of additional tests
.../testing/selftests/ftrace/test.d/functions | 6 +++--- .../test.d/kprobe/kprobe_syntax_errors.tc | 20 +++++++++---------- 2 files changed, 13 insertions(+), 13 deletions(-)
diff --git a/tools/testing/selftests/ftrace/test.d/functions b/tools/testing/selftests/ftrace/test.d/functions index 5d4550591ff9..b38c6eb029e8 100644 --- a/tools/testing/selftests/ftrace/test.d/functions +++ b/tools/testing/selftests/ftrace/test.d/functions @@ -114,11 +114,11 @@ yield() { }
ftrace_errlog_check() { # err-prefix command-with-error-pos-by-^ command-file - pos=$(echo -n "${2%^*}" | wc -c) # error position - command=$(echo "$2" | tr -d ^) + pos=$(printf "${2%^*}" | wc -c) # error position + command=$(printf "$2" | sed -e 's/^//g' -e 's/%/%%/g' -e 's/\/\\/g') echo "Test command: $command" echo > error_log - (! echo "$command" >> "$3" ) 2> /dev/null + (! printf "$command" >> "$3" ) 2> /dev/null grep "$1: error:" -A 3 error_log N=$(tail -n 1 error_log | wc -c) # " Command: " and "^\n" => 13 diff --git a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc index ef1e9bafb098..039c03d230b9 100644 --- a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc +++ b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc @@ -37,14 +37,14 @@ fi
check_error 'p vfs_read ^$none_var' # BAD_VAR
-check_error 'p vfs_read ^%none_reg' # BAD_REG_NAME +check_error 'p vfs_read ^%%none_reg' # BAD_REG_NAME check_error 'p vfs_read ^@12345678abcde' # BAD_MEM_ADDR check_error 'p vfs_read ^@+10' # FILE_ON_KPROBE
grep -q "imm-value" README && \ -check_error 'p vfs_read arg1=^x' # BAD_IMM +check_error 'p vfs_read arg1=\^x' # BAD_IMM grep -q "imm-string" README && \ -check_error 'p vfs_read arg1="abcd^' # IMMSTR_NO_CLOSE +check_error 'p vfs_read arg1=\"abcd^' # IMMSTR_NO_CLOSE
check_error 'p vfs_read ^+0@0)' # DEREF_NEED_BRACE check_error 'p vfs_read ^+0ab1(@0)' # BAD_DEREF_OFFS @@ -80,7 +80,7 @@ check_error 'p vfs_read arg1=^' # NO_ARG_BODY # instruction boundary check is valid on x86 (at this moment) case $(uname -m) in x86_64|i[3456]86) - echo 'p vfs_read' > kprobe_events + printf 'p vfs_read' > kprobe_events if grep -q FTRACE ../kprobes/list ; then check_error 'p ^vfs_read+3' # BAD_INSN_BNDRY (only if function-tracer is enabled) fi @@ -89,13 +89,13 @@ esac
# multiprobe errors if grep -q "Create/append/" README && grep -q "imm-value" README; then -echo 'p:kprobes/testevent _do_fork' > kprobe_events +printf 'p:kprobes/testevent _do_fork' > kprobe_events check_error '^r:kprobes/testevent do_exit' # DIFF_PROBE_TYPE -echo 'p:kprobes/testevent _do_fork abcd=\1' > kprobe_events -check_error 'p:kprobes/testevent _do_fork ^bcd=\1' # DIFF_ARG_TYPE -check_error 'p:kprobes/testevent _do_fork ^abcd=\1:u8' # DIFF_ARG_TYPE -check_error 'p:kprobes/testevent _do_fork ^abcd="foo"' # DIFF_ARG_TYPE -check_error '^p:kprobes/testevent _do_fork abcd=\1' # SAME_PROBE +printf 'p:kprobes/testevent _do_fork abcd=\1' > kprobe_events +check_error 'p:kprobes/testevent _do_fork ^bcd=\1' # DIFF_ARG_TYPE +check_error 'p:kprobes/testevent _do_fork ^abcd=\1:u8' # DIFF_ARG_TYPE +check_error 'p:kprobes/testevent _do_fork ^abcd=\"foo"'# DIFF_ARG_TYPE +check_error '^p:kprobes/testevent _do_fork abcd=\1' # SAME_PROBE fi
exit 0
On Wed, Mar 04, 2020 at 04:20:09PM -0600, Seth Forshee wrote:
Test cases which use echo to write strings containing backslashes fail with some shells, as echo's treatment of backslashes in strings varies between shell implementations. Use printf instead, as it should behave consistently across different shells. This requires adjustments to the strings to escape \ and % characters. ftrace_errlog_check() must also re-escape these characters after processing them to remove ^ characters.
Signed-off-by: Seth Forshee seth.forshee@canonical.com
Ping. Someone just asked me about this patch, and I noticed that it hasn't been applied or received any feedback.
Changes in v2:
- Escape backslashes for a couple of additional tests
.../testing/selftests/ftrace/test.d/functions | 6 +++--- .../test.d/kprobe/kprobe_syntax_errors.tc | 20 +++++++++---------- 2 files changed, 13 insertions(+), 13 deletions(-)
diff --git a/tools/testing/selftests/ftrace/test.d/functions b/tools/testing/selftests/ftrace/test.d/functions index 5d4550591ff9..b38c6eb029e8 100644 --- a/tools/testing/selftests/ftrace/test.d/functions +++ b/tools/testing/selftests/ftrace/test.d/functions @@ -114,11 +114,11 @@ yield() { } ftrace_errlog_check() { # err-prefix command-with-error-pos-by-^ command-file
- pos=$(echo -n "${2%^*}" | wc -c) # error position
- command=$(echo "$2" | tr -d ^)
- pos=$(printf "${2%^*}" | wc -c) # error position
- command=$(printf "$2" | sed -e 's/^//g' -e 's/%/%%/g' -e 's/\/\\/g') echo "Test command: $command" echo > error_log
- (! echo "$command" >> "$3" ) 2> /dev/null
- (! printf "$command" >> "$3" ) 2> /dev/null grep "$1: error:" -A 3 error_log N=$(tail -n 1 error_log | wc -c) # " Command: " and "^\n" => 13
diff --git a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc index ef1e9bafb098..039c03d230b9 100644 --- a/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc +++ b/tools/testing/selftests/ftrace/test.d/kprobe/kprobe_syntax_errors.tc @@ -37,14 +37,14 @@ fi check_error 'p vfs_read ^$none_var' # BAD_VAR -check_error 'p vfs_read ^%none_reg' # BAD_REG_NAME +check_error 'p vfs_read ^%%none_reg' # BAD_REG_NAME check_error 'p vfs_read ^@12345678abcde' # BAD_MEM_ADDR check_error 'p vfs_read ^@+10' # FILE_ON_KPROBE grep -q "imm-value" README && \ -check_error 'p vfs_read arg1=^x' # BAD_IMM +check_error 'p vfs_read arg1=\^x' # BAD_IMM grep -q "imm-string" README && \ -check_error 'p vfs_read arg1="abcd^' # IMMSTR_NO_CLOSE +check_error 'p vfs_read arg1=\"abcd^' # IMMSTR_NO_CLOSE check_error 'p vfs_read ^+0@0)' # DEREF_NEED_BRACE check_error 'p vfs_read ^+0ab1(@0)' # BAD_DEREF_OFFS @@ -80,7 +80,7 @@ check_error 'p vfs_read arg1=^' # NO_ARG_BODY # instruction boundary check is valid on x86 (at this moment) case $(uname -m) in x86_64|i[3456]86)
- echo 'p vfs_read' > kprobe_events
- printf 'p vfs_read' > kprobe_events if grep -q FTRACE ../kprobes/list ; then check_error 'p ^vfs_read+3' # BAD_INSN_BNDRY (only if function-tracer is enabled) fi
@@ -89,13 +89,13 @@ esac # multiprobe errors if grep -q "Create/append/" README && grep -q "imm-value" README; then -echo 'p:kprobes/testevent _do_fork' > kprobe_events +printf 'p:kprobes/testevent _do_fork' > kprobe_events check_error '^r:kprobes/testevent do_exit' # DIFF_PROBE_TYPE -echo 'p:kprobes/testevent _do_fork abcd=\1' > kprobe_events -check_error 'p:kprobes/testevent _do_fork ^bcd=\1' # DIFF_ARG_TYPE -check_error 'p:kprobes/testevent _do_fork ^abcd=\1:u8' # DIFF_ARG_TYPE -check_error 'p:kprobes/testevent _do_fork ^abcd="foo"' # DIFF_ARG_TYPE -check_error '^p:kprobes/testevent _do_fork abcd=\1' # SAME_PROBE +printf 'p:kprobes/testevent _do_fork abcd=\1' > kprobe_events +check_error 'p:kprobes/testevent _do_fork ^bcd=\1' # DIFF_ARG_TYPE +check_error 'p:kprobes/testevent _do_fork ^abcd=\1:u8' # DIFF_ARG_TYPE +check_error 'p:kprobes/testevent _do_fork ^abcd=\"foo"'# DIFF_ARG_TYPE +check_error '^p:kprobes/testevent _do_fork abcd=\1' # SAME_PROBE fi exit 0 -- 2.25.0
On 5/29/20 2:37 PM, Seth Forshee wrote:
On Wed, Mar 04, 2020 at 04:20:09PM -0600, Seth Forshee wrote:
Test cases which use echo to write strings containing backslashes fail with some shells, as echo's treatment of backslashes in strings varies between shell implementations. Use printf instead, as it should behave consistently across different shells. This requires adjustments to the strings to escape \ and % characters. ftrace_errlog_check() must also re-escape these characters after processing them to remove ^ characters.
Signed-off-by: Seth Forshee seth.forshee@canonical.com
Ping. Someone just asked me about this patch, and I noticed that it hasn't been applied or received any feedback.
I pulled in this patch from Masami:
selftests/ftrace: Use printf for backslash included command https://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest.git/co...
Looks like a duplicate.
Seth, Is your patch still needed?
+ Steve and Masami
thanks, -- Shuah
On Fri, May 29, 2020 at 03:26:06PM -0600, Shuah Khan wrote:
On 5/29/20 2:37 PM, Seth Forshee wrote:
On Wed, Mar 04, 2020 at 04:20:09PM -0600, Seth Forshee wrote:
Test cases which use echo to write strings containing backslashes fail with some shells, as echo's treatment of backslashes in strings varies between shell implementations. Use printf instead, as it should behave consistently across different shells. This requires adjustments to the strings to escape \ and % characters. ftrace_errlog_check() must also re-escape these characters after processing them to remove ^ characters.
Signed-off-by: Seth Forshee seth.forshee@canonical.com
Ping. Someone just asked me about this patch, and I noticed that it hasn't been applied or received any feedback.
I pulled in this patch from Masami:
selftests/ftrace: Use printf for backslash included command https://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest.git/co...
Looks like a duplicate.
Seth, Is your patch still needed?
Nope, Masami's patch seems to fix the issues addressed by my patch. Thanks!
linux-kselftest-mirror@lists.linaro.org