* [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c
@ 2021-06-03 14:57 Kumar N, Bhuvanendra
2021-06-03 15:38 ` Simon Marchi
0 siblings, 1 reply; 5+ messages in thread
From: Kumar N, Bhuvanendra @ 2021-06-03 14:57 UTC (permalink / raw)
To: Simon Marchi, gdb-patches
Cc: George, Jini Susan, Achra, Nitika, Sharma, Alok Kumar, E,
Nagajyothi, Tomar, Sourabh Singh
[-- Attachment #1: Type: text/plain, Size: 6668 bytes --]
[AMD Official Use Only - Internal Distribution Only]
Hi Simon and all,
Could you please review this GDB patch. Patch is attached with this email and also inlined here. Thanks in advance
Problem Description:
Following 8 test points started to fail after the clang backend patch(https://reviews.llvm.org/D91734).
FAIL: gdb.base/foll-exec.exp: step through execlp call
FAIL: gdb.base/foll-exec.exp: step after execlp call
FAIL: gdb.base/foll-exec.exp: print execd-program/global_i (after execlp)
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execlp)
FAIL: gdb.base/foll-exec.exp: print follow-exec/local_k (after execlp)
FAIL: gdb.base/foll-exec.exp: step through execl call
FAIL: gdb.base/foll-exec.exp: step after execl call
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execl)
Sample gdb.base/foll-exec.c test file snippet :
. . .
42 execlp (prog, /* tbreak-execlp */
43 prog,
44 "execlp arg1 from foll-exec",
45 (char *) 0);
46
47 printf ("foll-exec is about to execl(execd-prog)...\n");
. . .
Resolution:
We had discussed 2 ways of handling this issue earlier, i.e.
1. Adding few extra "next" command during these multi-line function call
2. combine these multi-line function call into single line in foll-exec.c
This PR has taken the latter approach and converted the multi-line function call into single line in foll-exec.c, thereby there is no change in .debug_line entries now and test case works as expected.
NOTE: earlier I had sent a test patch(sent on 2021-04-15) with first approach and it was suggested to follow the second approach, hence now I am sending this PR with the second approach
regards,
bhuvan
Patch content inlined:
From 77e09b9b9a19dfb35a748fd57d7a16eb52f1b6ee Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E2=80=9Cbhkumarn=E2=80=9D?= Bhuvanendra.KumarN@amd.com<mailto:Bhuvanendra.KumarN@amd.com>
Date: Thu, 3 Jun 2021 17:50:28 +0530
Subject: [PATCH] [gdb.base] Convert multi-line function call into
single line.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
After this clang backend patch(https://reviews.llvm.org/D91734), 8 test points
started to FAIL in this test case. As mentioned in this PR, "...this test is
trying to "next" over a function call; gcc attributes all parameter evaluation
to the function name, while clang will attribute each parameter to its own
location. And when the parameters span across multiple source lines, the
is_stmt heuristic kicks in, so we stop on each line with actual parameters...".
gdb.base/foll-exec.c test file snippet :
. . .
42 execlp (prog, /* tbreak-execlp */
43 prog,
44 "execlp arg1 from foll-exec",
45 (char *) 0);
46
47 printf ("foll-exec is about to execl(execd-prog)...\n");
. . .
Line table: (before clang backend patch for the above code snippet) :
0x000000b0: 84 address += 8, line += 2
0x000000000020196a 42 3 1 0 0
0x000000b1: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000b2: 41 address += 3, line += 5
0x000000000020197e 47 3 1 0 0
Line table: (after clang backend patch for the above code snippet) :
0x000000b5: 84 address += 8, line += 2
0x0000000000201958 42 11 1 0 0
0x000000b6: 05 DW_LNS_set_column (4)
0x000000b8: 75 address += 7, line += 1
0x000000000020195f 43 4 1 0 0
0x000000b9: 05 DW_LNS_set_column (3)
0x000000bb: 73 address += 7, line += -1
0x0000000000201966 42 3 1 0 0
0x000000bc: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000bd: 4f address += 4, line += 5
0x000000000020197b 47 3 1 0 0
Following 8 test points started to fail after the above clang backend patch.
FAIL: gdb.base/foll-exec.exp: step through execlp call
FAIL: gdb.base/foll-exec.exp: step after execlp call
FAIL: gdb.base/foll-exec.exp: print execd-program/global_i (after execlp)
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execlp)
FAIL: gdb.base/foll-exec.exp: print follow-exec/local_k (after execlp)
FAIL: gdb.base/foll-exec.exp: step through execl call
FAIL: gdb.base/foll-exec.exp: step after execl call
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execl)
As we can note, reason for these new test failures is due to additional
.debug_line entries getting created in case of clang compiler, hence to fix
this issue, test case required either additional "next" command during
these multi-line function call or combine these multi-line function call into
single line. This PR has taken the latter approach and converted the multi-line
function call into single line in foll-exec.c, thereby there is no change in
.debug_line entries now and test case works as expected.
---
gdb/testsuite/ChangeLog | 5 +++++
gdb/testsuite/gdb.base/foll-exec.c | 11 ++---------
2 files changed, 7 insertions(+), 9 deletions(-)
diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog
index 7959f58c3c4..1873f9ddf05 100644
--- a/gdb/testsuite/ChangeLog
+++ b/gdb/testsuite/ChangeLog
@@ -1,3 +1,8 @@
+2021-06-03 Bhuvanendra Kumar N Bhuvanendra.KumarN@amd.com<mailto:Bhuvanendra.KumarN@amd.com>
+
+ * gdb.base/foll-exec.c: convert the multi-line function call into
+ single line.
+
2021-06-02 Bernd Edlinger bernd.edlinger@hotmail.de<mailto:bernd.edlinger@hotmail.de>
* gdb.dwarf2/per-bfd-sharing.exp: Fix temp-dir leakage.
diff --git a/gdb/testsuite/gdb.base/foll-exec.c b/gdb/testsuite/gdb.base/foll-exec.c
index fea62b06db4..f1b97aca4a4 100644
--- a/gdb/testsuite/gdb.base/foll-exec.c
+++ b/gdb/testsuite/gdb.base/foll-exec.c
@@ -39,18 +39,11 @@ int main (int argc, char ** argv)
memcpy (prog + len - 9, "execd-prog", 10);
prog[len + 1] = 0;
- execlp (prog, /* tbreak-execlp */
- prog,
- "execlp arg1 from foll-exec",
- (char *) 0);
+ execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from foll-exec", (char *) 0);
printf ("foll-exec is about to execl(execd-prog)...\n");
- execl (prog, /* tbreak-execl */
- prog,
- "execl arg1 from foll-exec",
- "execl arg2 from foll-exec",
- (char *) 0);
+ execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec", "execl arg2 from foll-exec", (char *) 0);
{
static char * argv[] = {
--
2.17.1
[-- Attachment #2: gdb.base-Convert-multi-line-function-call-into.patch --]
[-- Type: application/octet-stream, Size: 4726 bytes --]
From 77e09b9b9a19dfb35a748fd57d7a16eb52f1b6ee Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E2=80=9Cbhkumarn=E2=80=9D?= <Bhuvanendra.KumarN@amd.com>
Date: Thu, 3 Jun 2021 17:50:28 +0530
Subject: [PATCH] [gdb.base] Convert multi-line function call into
single line.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
After this clang backend patch(https://reviews.llvm.org/D91734), 8 test points
started to FAIL in this test case. As mentioned in this PR, "...this test is
trying to "next" over a function call; gcc attributes all parameter evaluation
to the function name, while clang will attribute each parameter to its own
location. And when the parameters span across multiple source lines, the
is_stmt heuristic kicks in, so we stop on each line with actual parameters...".
gdb.base/foll-exec.c test file snippet :
. . .
42 execlp (prog, /* tbreak-execlp */
43 prog,
44 "execlp arg1 from foll-exec",
45 (char *) 0);
46
47 printf ("foll-exec is about to execl(execd-prog)...\n");
. . .
Line table: (before clang backend patch for the above code snippet) :
0x000000b0: 84 address += 8, line += 2
0x000000000020196a 42 3 1 0 0
0x000000b1: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000b2: 41 address += 3, line += 5
0x000000000020197e 47 3 1 0 0
Line table: (after clang backend patch for the above code snippet) :
0x000000b5: 84 address += 8, line += 2
0x0000000000201958 42 11 1 0 0
0x000000b6: 05 DW_LNS_set_column (4)
0x000000b8: 75 address += 7, line += 1
0x000000000020195f 43 4 1 0 0
0x000000b9: 05 DW_LNS_set_column (3)
0x000000bb: 73 address += 7, line += -1
0x0000000000201966 42 3 1 0 0
0x000000bc: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000bd: 4f address += 4, line += 5
0x000000000020197b 47 3 1 0 0
Following 8 test points started to fail after the above clang backend patch.
FAIL: gdb.base/foll-exec.exp: step through execlp call
FAIL: gdb.base/foll-exec.exp: step after execlp call
FAIL: gdb.base/foll-exec.exp: print execd-program/global_i (after execlp)
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execlp)
FAIL: gdb.base/foll-exec.exp: print follow-exec/local_k (after execlp)
FAIL: gdb.base/foll-exec.exp: step through execl call
FAIL: gdb.base/foll-exec.exp: step after execl call
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execl)
As we can note, reason for these new test failures is due to additional
.debug_line entries getting created in case of clang compiler, hence to fix
this issue, test case required either additional “next” command during
these multi-line function call or combine these multi-line function call into
single line. This PR has taken the latter approach and converted the multi-line
function call into single line in foll-exec.c, thereby there is no change in
.debug_line entries now and test case works as expected.
---
gdb/testsuite/ChangeLog | 5 +++++
gdb/testsuite/gdb.base/foll-exec.c | 11 ++---------
2 files changed, 7 insertions(+), 9 deletions(-)
diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog
index 7959f58c3c4..1873f9ddf05 100644
--- a/gdb/testsuite/ChangeLog
+++ b/gdb/testsuite/ChangeLog
@@ -1,3 +1,8 @@
+2021-06-03 Bhuvanendra Kumar N <Bhuvanendra.KumarN@amd.com>
+
+ * gdb.base/foll-exec.c: convert the multi-line function call into
+ single line.
+
2021-06-02 Bernd Edlinger <bernd.edlinger@hotmail.de>
* gdb.dwarf2/per-bfd-sharing.exp: Fix temp-dir leakage.
diff --git a/gdb/testsuite/gdb.base/foll-exec.c b/gdb/testsuite/gdb.base/foll-exec.c
index fea62b06db4..f1b97aca4a4 100644
--- a/gdb/testsuite/gdb.base/foll-exec.c
+++ b/gdb/testsuite/gdb.base/foll-exec.c
@@ -39,18 +39,11 @@ int main (int argc, char ** argv)
memcpy (prog + len - 9, "execd-prog", 10);
prog[len + 1] = 0;
- execlp (prog, /* tbreak-execlp */
- prog,
- "execlp arg1 from foll-exec",
- (char *) 0);
+ execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from foll-exec", (char *) 0);
printf ("foll-exec is about to execl(execd-prog)...\n");
- execl (prog, /* tbreak-execl */
- prog,
- "execl arg1 from foll-exec",
- "execl arg2 from foll-exec",
- (char *) 0);
+ execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec", "execl arg2 from foll-exec", (char *) 0);
{
static char * argv[] = {
--
2.17.1
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c
2021-06-03 14:57 [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c Kumar N, Bhuvanendra
@ 2021-06-03 15:38 ` Simon Marchi
2021-06-07 6:31 ` Kumar N, Bhuvanendra
0 siblings, 1 reply; 5+ messages in thread
From: Simon Marchi @ 2021-06-03 15:38 UTC (permalink / raw)
To: Kumar N, Bhuvanendra, gdb-patches
Cc: George, Jini Susan, Achra, Nitika, Sharma, Alok Kumar, E,
Nagajyothi, Tomar, Sourabh Singh
> diff --git a/gdb/testsuite/gdb.base/foll-exec.c b/gdb/testsuite/gdb.base/foll-exec.c
>
> index fea62b06db4..f1b97aca4a4 100644
>
> --- a/gdb/testsuite/gdb.base/foll-exec.c
>
> +++ b/gdb/testsuite/gdb.base/foll-exec.c
>
> @@ -39,18 +39,11 @@ int main (int argc, char ** argv)
>
> memcpy (prog + len - 9, "execd-prog", 10);
>
> prog[len + 1] = 0;
>
> - execlp (prog, /* tbreak-execlp */
>
> - prog,
>
> - "execlp arg1 from foll-exec",
>
> - (char *) 0);
>
> + execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from foll-exec", (char *) 0);
>
> printf ("foll-exec is about to execl(execd-prog)...\n");
>
> - execl (prog, /* tbreak-execl */
>
> - prog,
>
> - "execl arg1 from foll-exec",
>
> - "execl arg2 from foll-exec",
>
> - (char *) 0);
>
> + execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec", "execl arg2 from foll-exec", (char *) 0);
>
> {
>
> static char * argv[] = {
Since these lines exceed the normal line length limit of 80 columns,
please add a comment above to say why the call is all on one line.
Otherwise, someone may be tempted to "fix it" and put it back on
multiple lines (such a comment may be useful in the other test too).
Thanks,
Simon
^ permalink raw reply [flat|nested] 5+ messages in thread
* RE: [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c
2021-06-03 15:38 ` Simon Marchi
@ 2021-06-07 6:31 ` Kumar N, Bhuvanendra
2021-06-07 15:07 ` Simon Marchi
0 siblings, 1 reply; 5+ messages in thread
From: Kumar N, Bhuvanendra @ 2021-06-07 6:31 UTC (permalink / raw)
To: Simon Marchi, gdb-patches
Cc: George, Jini Susan, Achra, Nitika, Sharma, Alok Kumar, E,
Nagajyothi, Tomar, Sourabh Singh
[-- Attachment #1: Type: text/plain, Size: 7537 bytes --]
[AMD Official Use Only - Internal Distribution Only]
Hi Simon,
> Since these lines exceed the normal line length limit of 80 columns, please add a comment above to say why the call is all on one line.
I have added the required comment in the source file, could you please review/approve the updated patch.
I have attached the updated patch and also inlined here
regards,
bhuvan
Patch inlined:
From 8f77e54fad98f17f3f34d07a0275de521aeed74f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E2=80=9Cbhkumarn=E2=80=9D?= <Bhuvanendra.KumarN@amd.com>
Date: Thu, 3 Jun 2021 17:50:28 +0530
Subject: [PATCH] [gdb.base] Convert multi-line function call into
single line.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
After this clang backend patch(https://reviews.llvm.org/D91734), 8 test points
started to FAIL in this test case. As mentioned in this PR, "...this test is
trying to "next" over a function call; gcc attributes all parameter evaluation
to the function name, while clang will attribute each parameter to its own
location. And when the parameters span across multiple source lines, the
is_stmt heuristic kicks in, so we stop on each line with actual parameters...".
gdb.base/foll-exec.c test file snippet :
. . .
42 execlp (prog, /* tbreak-execlp */
43 prog,
44 "execlp arg1 from foll-exec",
45 (char *) 0);
46
47 printf ("foll-exec is about to execl(execd-prog)...\n");
. . .
Line table: (before clang backend patch for the above code snippet) :
0x000000b0: 84 address += 8, line += 2
0x000000000020196a 42 3 1 0 0
0x000000b1: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000b2: 41 address += 3, line += 5
0x000000000020197e 47 3 1 0 0
Line table: (after clang backend patch for the above code snippet) :
0x000000b5: 84 address += 8, line += 2
0x0000000000201958 42 11 1 0 0
0x000000b6: 05 DW_LNS_set_column (4)
0x000000b8: 75 address += 7, line += 1
0x000000000020195f 43 4 1 0 0
0x000000b9: 05 DW_LNS_set_column (3)
0x000000bb: 73 address += 7, line += -1
0x0000000000201966 42 3 1 0 0
0x000000bc: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000bd: 4f address += 4, line += 5
0x000000000020197b 47 3 1 0 0
Following 8 test points started to fail after the above clang backend patch.
FAIL: gdb.base/foll-exec.exp: step through execlp call
FAIL: gdb.base/foll-exec.exp: step after execlp call
FAIL: gdb.base/foll-exec.exp: print execd-program/global_i (after execlp)
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execlp)
FAIL: gdb.base/foll-exec.exp: print follow-exec/local_k (after execlp)
FAIL: gdb.base/foll-exec.exp: step through execl call
FAIL: gdb.base/foll-exec.exp: step after execl call
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execl)
As we can note, reason for these new test failures is due to additional
.debug_line entries getting created in case of clang compiler, hence to fix
this issue, test case required either additional “next†command during
these multi-line function call or combine these multi-line function call into
single line. This PR has taken the latter approach and converted the multi-line
function call into single line in foll-exec.c, thereby there is no change in
.debug_line entries now and test case works as expected.
---
gdb/testsuite/ChangeLog | 5 +++++
gdb/testsuite/gdb.base/foll-exec.c | 19 ++++++++++---------
2 files changed, 15 insertions(+), 9 deletions(-)
diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog
index 7959f58c3c4..1873f9ddf05 100644
--- a/gdb/testsuite/ChangeLog
+++ b/gdb/testsuite/ChangeLog
@@ -1,3 +1,8 @@
+2021-06-03 Bhuvanendra Kumar N <Bhuvanendra.KumarN@amd.com>
+
+ * gdb.base/foll-exec.c: convert the multi-line function call into
+ single line.
+
2021-06-02 Bernd Edlinger <bernd.edlinger@hotmail.de>
* gdb.dwarf2/per-bfd-sharing.exp: Fix temp-dir leakage.
diff --git a/gdb/testsuite/gdb.base/foll-exec.c b/gdb/testsuite/gdb.base/foll-exec.c
index fea62b06db4..77a29860ebc 100644
--- a/gdb/testsuite/gdb.base/foll-exec.c
+++ b/gdb/testsuite/gdb.base/foll-exec.c
@@ -39,18 +39,19 @@ int main (int argc, char ** argv)
memcpy (prog + len - 9, "execd-prog", 10);
prog[len + 1] = 0;
- execlp (prog, /* tbreak-execlp */
- prog,
- "execlp arg1 from foll-exec",
- (char *) 0);
+ /* In the following function call, maximum line length exceed the limit 80.
+ This is intentional and required for clang compiler such that complete
+ function call should be in a single line, please do not make it
+ multi-line */
+ execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from foll-exec", (char *) 0);
printf ("foll-exec is about to execl(execd-prog)...\n");
- execl (prog, /* tbreak-execl */
- prog,
- "execl arg1 from foll-exec",
- "execl arg2 from foll-exec",
- (char *) 0);
+ /* In the following function call, maximum line length exceed the limit 80.
+ This is intentional and required for clang compiler such that complete
+ function call should be in a single line, please do not make it
+ multi-line */
+ execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec", "execl arg2 from foll-exec", (char *) 0);
{
static char * argv[] = {
--
2.17.1
-----Original Message-----
From: Simon Marchi <simon.marchi@polymtl.ca>
Sent: Thursday, June 3, 2021 9:09 PM
To: Kumar N, Bhuvanendra <Bhuvanendra.KumarN@amd.com>; gdb-patches@sourceware.org
Cc: George, Jini Susan <JiniSusan.George@amd.com>; Achra, Nitika <Nitika.Achra@amd.com>; Sharma, Alok Kumar <AlokKumar.Sharma@amd.com>; E, Nagajyothi <Nagajyothi.E@amd.com>; Tomar, Sourabh Singh <SourabhSingh.Tomar@amd.com>
Subject: Re: [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c
[CAUTION: External Email]
> diff --git a/gdb/testsuite/gdb.base/foll-exec.c
> b/gdb/testsuite/gdb.base/foll-exec.c
>
> index fea62b06db4..f1b97aca4a4 100644
>
> --- a/gdb/testsuite/gdb.base/foll-exec.c
>
> +++ b/gdb/testsuite/gdb.base/foll-exec.c
>
> @@ -39,18 +39,11 @@ int main (int argc, char ** argv)
>
> memcpy (prog + len - 9, "execd-prog", 10);
>
> prog[len + 1] = 0;
>
> - execlp (prog, /* tbreak-execlp */
>
> - prog,
>
> - "execlp arg1 from foll-exec",
>
> - (char *) 0);
>
> + execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from
> + foll-exec", (char *) 0);
>
> printf ("foll-exec is about to execl(execd-prog)...\n");
>
> - execl (prog, /* tbreak-execl */
>
> - prog,
>
> - "execl arg1 from foll-exec",
>
> - "execl arg2 from foll-exec",
>
> - (char *) 0);
>
> + execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec",
> + "execl arg2 from foll-exec", (char *) 0);
>
> {
>
> static char * argv[] = {
Since these lines exceed the normal line length limit of 80 columns, please add a comment above to say why the call is all on one line.
Otherwise, someone may be tempted to "fix it" and put it back on multiple lines (such a comment may be useful in the other test too).
Thanks,
Simon
[-- Attachment #2: gdb.base-Convert-multi-line-function-call-into.patch --]
[-- Type: application/octet-stream, Size: 5117 bytes --]
From 8f77e54fad98f17f3f34d07a0275de521aeed74f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E2=80=9Cbhkumarn=E2=80=9D?= <Bhuvanendra.KumarN@amd.com>
Date: Thu, 3 Jun 2021 17:50:28 +0530
Subject: [PATCH] [gdb.base] Convert multi-line function call into
single line.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
After this clang backend patch(https://reviews.llvm.org/D91734), 8 test points
started to FAIL in this test case. As mentioned in this PR, "...this test is
trying to "next" over a function call; gcc attributes all parameter evaluation
to the function name, while clang will attribute each parameter to its own
location. And when the parameters span across multiple source lines, the
is_stmt heuristic kicks in, so we stop on each line with actual parameters...".
gdb.base/foll-exec.c test file snippet :
. . .
42 execlp (prog, /* tbreak-execlp */
43 prog,
44 "execlp arg1 from foll-exec",
45 (char *) 0);
46
47 printf ("foll-exec is about to execl(execd-prog)...\n");
. . .
Line table: (before clang backend patch for the above code snippet) :
0x000000b0: 84 address += 8, line += 2
0x000000000020196a 42 3 1 0 0
0x000000b1: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000b2: 41 address += 3, line += 5
0x000000000020197e 47 3 1 0 0
Line table: (after clang backend patch for the above code snippet) :
0x000000b5: 84 address += 8, line += 2
0x0000000000201958 42 11 1 0 0
0x000000b6: 05 DW_LNS_set_column (4)
0x000000b8: 75 address += 7, line += 1
0x000000000020195f 43 4 1 0 0
0x000000b9: 05 DW_LNS_set_column (3)
0x000000bb: 73 address += 7, line += -1
0x0000000000201966 42 3 1 0 0
0x000000bc: 08 DW_LNS_const_add_pc (0x0000000000000011)
0x000000bd: 4f address += 4, line += 5
0x000000000020197b 47 3 1 0 0
Following 8 test points started to fail after the above clang backend patch.
FAIL: gdb.base/foll-exec.exp: step through execlp call
FAIL: gdb.base/foll-exec.exp: step after execlp call
FAIL: gdb.base/foll-exec.exp: print execd-program/global_i (after execlp)
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execlp)
FAIL: gdb.base/foll-exec.exp: print follow-exec/local_k (after execlp)
FAIL: gdb.base/foll-exec.exp: step through execl call
FAIL: gdb.base/foll-exec.exp: step after execl call
FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execl)
As we can note, reason for these new test failures is due to additional
.debug_line entries getting created in case of clang compiler, hence to fix
this issue, test case required either additional “next” command during
these multi-line function call or combine these multi-line function call into
single line. This PR has taken the latter approach and converted the multi-line
function call into single line in foll-exec.c, thereby there is no change in
.debug_line entries now and test case works as expected.
---
gdb/testsuite/ChangeLog | 5 +++++
gdb/testsuite/gdb.base/foll-exec.c | 19 ++++++++++---------
2 files changed, 15 insertions(+), 9 deletions(-)
diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog
index 7959f58c3c4..1873f9ddf05 100644
--- a/gdb/testsuite/ChangeLog
+++ b/gdb/testsuite/ChangeLog
@@ -1,3 +1,8 @@
+2021-06-03 Bhuvanendra Kumar N <Bhuvanendra.KumarN@amd.com>
+
+ * gdb.base/foll-exec.c: convert the multi-line function call into
+ single line.
+
2021-06-02 Bernd Edlinger <bernd.edlinger@hotmail.de>
* gdb.dwarf2/per-bfd-sharing.exp: Fix temp-dir leakage.
diff --git a/gdb/testsuite/gdb.base/foll-exec.c b/gdb/testsuite/gdb.base/foll-exec.c
index fea62b06db4..77a29860ebc 100644
--- a/gdb/testsuite/gdb.base/foll-exec.c
+++ b/gdb/testsuite/gdb.base/foll-exec.c
@@ -39,18 +39,19 @@ int main (int argc, char ** argv)
memcpy (prog + len - 9, "execd-prog", 10);
prog[len + 1] = 0;
- execlp (prog, /* tbreak-execlp */
- prog,
- "execlp arg1 from foll-exec",
- (char *) 0);
+ /* In the following function call, maximum line length exceed the limit 80.
+ This is intentional and required for clang compiler such that complete
+ function call should be in a single line, please do not make it
+ multi-line */
+ execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from foll-exec", (char *) 0);
printf ("foll-exec is about to execl(execd-prog)...\n");
- execl (prog, /* tbreak-execl */
- prog,
- "execl arg1 from foll-exec",
- "execl arg2 from foll-exec",
- (char *) 0);
+ /* In the following function call, maximum line length exceed the limit 80.
+ This is intentional and required for clang compiler such that complete
+ function call should be in a single line, please do not make it
+ multi-line */
+ execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec", "execl arg2 from foll-exec", (char *) 0);
{
static char * argv[] = {
--
2.17.1
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c
2021-06-07 6:31 ` Kumar N, Bhuvanendra
@ 2021-06-07 15:07 ` Simon Marchi
2021-06-07 15:19 ` Kumar N, Bhuvanendra
0 siblings, 1 reply; 5+ messages in thread
From: Simon Marchi @ 2021-06-07 15:07 UTC (permalink / raw)
To: Kumar N, Bhuvanendra, gdb-patches
Cc: George, Jini Susan, Achra, Nitika, Sharma, Alok Kumar, E,
Nagajyothi, Tomar, Sourabh Singh
On 2021-06-07 2:31 a.m., Kumar N, Bhuvanendra wrote:
> [AMD Official Use Only - Internal Distribution Only]
>
> Hi Simon,
>
>> Since these lines exceed the normal line length limit of 80 columns, please add a comment above to say why the call is all on one line.
>
> I have added the required comment in the source file, could you please review/approve the updated patch.
> I have attached the updated patch and also inlined here
>
> regards,
> bhuvan
>
> Patch inlined:
>
> From 8f77e54fad98f17f3f34d07a0275de521aeed74f Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?=E2=80=9Cbhkumarn=E2=80=9D?= <Bhuvanendra.KumarN@amd.com>
> Date: Thu, 3 Jun 2021 17:50:28 +0530
> Subject: [PATCH] [gdb.base] Convert multi-line function call into
> single line.
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
>
> After this clang backend patch(https://reviews.llvm.org/D91734), 8 test points
> started to FAIL in this test case. As mentioned in this PR, "...this test is
> trying to "next" over a function call; gcc attributes all parameter evaluation
> to the function name, while clang will attribute each parameter to its own
> location. And when the parameters span across multiple source lines, the
> is_stmt heuristic kicks in, so we stop on each line with actual parameters...".
>
> gdb.base/foll-exec.c test file snippet :
> . . .
> 42 execlp (prog, /* tbreak-execlp */
> 43 prog,
> 44 "execlp arg1 from foll-exec",
> 45 (char *) 0);
> 46
> 47 printf ("foll-exec is about to execl(execd-prog)...\n");
> . . .
>
> Line table: (before clang backend patch for the above code snippet) :
> 0x000000b0: 84 address += 8, line += 2
> 0x000000000020196a 42 3 1 0 0
> 0x000000b1: 08 DW_LNS_const_add_pc (0x0000000000000011)
> 0x000000b2: 41 address += 3, line += 5
> 0x000000000020197e 47 3 1 0 0
>
> Line table: (after clang backend patch for the above code snippet) :
> 0x000000b5: 84 address += 8, line += 2
> 0x0000000000201958 42 11 1 0 0
> 0x000000b6: 05 DW_LNS_set_column (4)
> 0x000000b8: 75 address += 7, line += 1
> 0x000000000020195f 43 4 1 0 0
> 0x000000b9: 05 DW_LNS_set_column (3)
> 0x000000bb: 73 address += 7, line += -1
> 0x0000000000201966 42 3 1 0 0
> 0x000000bc: 08 DW_LNS_const_add_pc (0x0000000000000011)
> 0x000000bd: 4f address += 4, line += 5
> 0x000000000020197b 47 3 1 0 0
>
> Following 8 test points started to fail after the above clang backend patch.
>
> FAIL: gdb.base/foll-exec.exp: step through execlp call
> FAIL: gdb.base/foll-exec.exp: step after execlp call
> FAIL: gdb.base/foll-exec.exp: print execd-program/global_i (after execlp)
> FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execlp)
> FAIL: gdb.base/foll-exec.exp: print follow-exec/local_k (after execlp)
> FAIL: gdb.base/foll-exec.exp: step through execl call
> FAIL: gdb.base/foll-exec.exp: step after execl call
> FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after execl)
>
> As we can note, reason for these new test failures is due to additional
> .debug_line entries getting created in case of clang compiler, hence to fix
> this issue, test case required either additional “next†command during
> these multi-line function call or combine these multi-line function call into
> single line. This PR has taken the latter approach and converted the multi-line
> function call into single line in foll-exec.c, thereby there is no change in
> .debug_line entries now and test case works as expected.
> ---
> gdb/testsuite/ChangeLog | 5 +++++
> gdb/testsuite/gdb.base/foll-exec.c | 19 ++++++++++---------
> 2 files changed, 15 insertions(+), 9 deletions(-)
>
> diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog
> index 7959f58c3c4..1873f9ddf05 100644
> --- a/gdb/testsuite/ChangeLog
> +++ b/gdb/testsuite/ChangeLog
> @@ -1,3 +1,8 @@
> +2021-06-03 Bhuvanendra Kumar N <Bhuvanendra.KumarN@amd.com>
> +
> + * gdb.base/foll-exec.c: convert the multi-line function call into
> + single line.
> +
> 2021-06-02 Bernd Edlinger <bernd.edlinger@hotmail.de>
>
> * gdb.dwarf2/per-bfd-sharing.exp: Fix temp-dir leakage.
> diff --git a/gdb/testsuite/gdb.base/foll-exec.c b/gdb/testsuite/gdb.base/foll-exec.c
> index fea62b06db4..77a29860ebc 100644
> --- a/gdb/testsuite/gdb.base/foll-exec.c
> +++ b/gdb/testsuite/gdb.base/foll-exec.c
> @@ -39,18 +39,19 @@ int main (int argc, char ** argv)
> memcpy (prog + len - 9, "execd-prog", 10);
> prog[len + 1] = 0;
>
> - execlp (prog, /* tbreak-execlp */
> - prog,
> - "execlp arg1 from foll-exec",
> - (char *) 0);
> + /* In the following function call, maximum line length exceed the limit 80.
> + This is intentional and required for clang compiler such that complete
> + function call should be in a single line, please do not make it
> + multi-line */
> + execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from foll-exec", (char *) 0);
>
> printf ("foll-exec is about to execl(execd-prog)...\n");
>
> - execl (prog, /* tbreak-execl */
> - prog,
> - "execl arg1 from foll-exec",
> - "execl arg2 from foll-exec",
> - (char *) 0);
> + /* In the following function call, maximum line length exceed the limit 80.
> + This is intentional and required for clang compiler such that complete
> + function call should be in a single line, please do not make it
> + multi-line */
> + execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec", "execl arg2 from foll-exec", (char *) 0);
>
> {
> static char * argv[] = {
>
This is ok, please just make sure to finish the sentence in the comment
appropriately, with a period and two spaces before the */.
Simon
^ permalink raw reply [flat|nested] 5+ messages in thread
* RE: [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c
2021-06-07 15:07 ` Simon Marchi
@ 2021-06-07 15:19 ` Kumar N, Bhuvanendra
0 siblings, 0 replies; 5+ messages in thread
From: Kumar N, Bhuvanendra @ 2021-06-07 15:19 UTC (permalink / raw)
To: Simon Marchi, gdb-patches
Cc: George, Jini Susan, Achra, Nitika, Sharma, Alok Kumar, E,
Nagajyothi, Tomar, Sourabh Singh
[AMD Official Use Only - Internal Distribution Only]
Hi,
> please just make sure to finish the sentence in the comment appropriately, with a period and two spaces before the */.
Sure, will make these changes before pushing the changes, Thanks Simon for your review comments.
regards,
bhuvan
-----Original Message-----
From: Simon Marchi <simon.marchi@polymtl.ca>
Sent: Monday, June 7, 2021 8:37 PM
To: Kumar N, Bhuvanendra <Bhuvanendra.KumarN@amd.com>; gdb-patches@sourceware.org
Cc: George, Jini Susan <JiniSusan.George@amd.com>; Achra, Nitika <Nitika.Achra@amd.com>; Sharma, Alok Kumar <AlokKumar.Sharma@amd.com>; E, Nagajyothi <Nagajyothi.E@amd.com>; Tomar, Sourabh Singh <SourabhSingh.Tomar@amd.com>
Subject: Re: [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c
[CAUTION: External Email]
On 2021-06-07 2:31 a.m., Kumar N, Bhuvanendra wrote:
> [AMD Official Use Only - Internal Distribution Only]
>
> Hi Simon,
>
>> Since these lines exceed the normal line length limit of 80 columns, please add a comment above to say why the call is all on one line.
>
> I have added the required comment in the source file, could you please review/approve the updated patch.
> I have attached the updated patch and also inlined here
>
> regards,
> bhuvan
>
> Patch inlined:
>
> From 8f77e54fad98f17f3f34d07a0275de521aeed74f Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?=E2=80=9Cbhkumarn=E2=80=9D?=
> <Bhuvanendra.KumarN@amd.com>
> Date: Thu, 3 Jun 2021 17:50:28 +0530
> Subject: [PATCH] [gdb.base] Convert multi-line function call into
> single line.
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
>
> After this clang backend
> patch(https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%
> 2Freviews.llvm.org%2FD91734&data=04%7C01%7CBhuvanendra.KumarN%40am
> d.com%7C448d6971a0fb40f9a9c408d929c5f931%7C3dd8961fe4884e608e11a82d994
> e183d%7C0%7C0%7C637586752539681728%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4
> wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=FfdJKw05LGNtRgJMPTTNsKXKIC86roCe6zHSwoQr%2FUA%3D&reserved=0), 8 test points started to FAIL in this test case. As mentioned in this PR, "...this test is trying to "next" over a function call; gcc attributes all parameter evaluation to the function name, while clang will attribute each parameter to its own location. And when the parameters span across multiple source lines, the is_stmt heuristic kicks in, so we stop on each line with actual parameters...".
>
> gdb.base/foll-exec.c test file snippet :
> . . .
> 42 execlp (prog, /* tbreak-execlp */
> 43 prog,
> 44 "execlp arg1 from foll-exec",
> 45 (char *) 0);
> 46
> 47 printf ("foll-exec is about to execl(execd-prog)...\n");
> . . .
>
> Line table: (before clang backend patch for the above code snippet) :
> 0x000000b0: 84 address += 8, line += 2
> 0x000000000020196a 42 3 1 0 0
> 0x000000b1: 08 DW_LNS_const_add_pc (0x0000000000000011)
> 0x000000b2: 41 address += 3, line += 5
> 0x000000000020197e 47 3 1 0 0
>
> Line table: (after clang backend patch for the above code snippet) :
> 0x000000b5: 84 address += 8, line += 2
> 0x0000000000201958 42 11 1 0 0
> 0x000000b6: 05 DW_LNS_set_column (4)
> 0x000000b8: 75 address += 7, line += 1
> 0x000000000020195f 43 4 1 0 0
> 0x000000b9: 05 DW_LNS_set_column (3)
> 0x000000bb: 73 address += 7, line += -1
> 0x0000000000201966 42 3 1 0 0
> 0x000000bc: 08 DW_LNS_const_add_pc (0x0000000000000011)
> 0x000000bd: 4f address += 4, line += 5
> 0x000000000020197b 47 3 1 0 0
>
> Following 8 test points started to fail after the above clang backend patch.
>
> FAIL: gdb.base/foll-exec.exp: step through execlp call
> FAIL: gdb.base/foll-exec.exp: step after execlp call
> FAIL: gdb.base/foll-exec.exp: print execd-program/global_i (after
> execlp)
> FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after
> execlp)
> FAIL: gdb.base/foll-exec.exp: print follow-exec/local_k (after execlp)
> FAIL: gdb.base/foll-exec.exp: step through execl call
> FAIL: gdb.base/foll-exec.exp: step after execl call
> FAIL: gdb.base/foll-exec.exp: print execd-program/local_j (after
> execl)
>
> As we can note, reason for these new test failures is due to
> additional .debug_line entries getting created in case of clang
> compiler, hence to fix this issue, test case required either
> additional “next†command during these multi-line function call or
> combine these multi-line function call into single line. This PR has
> taken the latter approach and converted the multi-line function call
> into single line in foll-exec.c, thereby there is no change in .debug_line entries now and test case works as expected.
> ---
> gdb/testsuite/ChangeLog | 5 +++++
> gdb/testsuite/gdb.base/foll-exec.c | 19 ++++++++++---------
> 2 files changed, 15 insertions(+), 9 deletions(-)
>
> diff --git a/gdb/testsuite/ChangeLog b/gdb/testsuite/ChangeLog index
> 7959f58c3c4..1873f9ddf05 100644
> --- a/gdb/testsuite/ChangeLog
> +++ b/gdb/testsuite/ChangeLog
> @@ -1,3 +1,8 @@
> +2021-06-03 Bhuvanendra Kumar N <Bhuvanendra.KumarN@amd.com>
> +
> + * gdb.base/foll-exec.c: convert the multi-line function call into
> + single line.
> +
> 2021-06-02 Bernd Edlinger <bernd.edlinger@hotmail.de>
>
> * gdb.dwarf2/per-bfd-sharing.exp: Fix temp-dir leakage.
> diff --git a/gdb/testsuite/gdb.base/foll-exec.c
> b/gdb/testsuite/gdb.base/foll-exec.c
> index fea62b06db4..77a29860ebc 100644
> --- a/gdb/testsuite/gdb.base/foll-exec.c
> +++ b/gdb/testsuite/gdb.base/foll-exec.c
> @@ -39,18 +39,19 @@ int main (int argc, char ** argv)
> memcpy (prog + len - 9, "execd-prog", 10);
> prog[len + 1] = 0;
>
> - execlp (prog, /* tbreak-execlp */
> - prog,
> - "execlp arg1 from foll-exec",
> - (char *) 0);
> + /* In the following function call, maximum line length exceed the limit 80.
> + This is intentional and required for clang compiler such that complete
> + function call should be in a single line, please do not make it
> + multi-line */
> + execlp (prog, /* tbreak-execlp */ prog, "execlp arg1 from
> + foll-exec", (char *) 0);
>
> printf ("foll-exec is about to execl(execd-prog)...\n");
>
> - execl (prog, /* tbreak-execl */
> - prog,
> - "execl arg1 from foll-exec",
> - "execl arg2 from foll-exec",
> - (char *) 0);
> + /* In the following function call, maximum line length exceed the limit 80.
> + This is intentional and required for clang compiler such that complete
> + function call should be in a single line, please do not make it
> + multi-line */
> + execl (prog, /* tbreak-execl */ prog, "execl arg1 from foll-exec", "execl arg2 from foll-exec", (char *) 0);
>
> {
> static char * argv[] = {
>
This is ok, please just make sure to finish the sentence in the comment appropriately, with a period and two spaces before the */.
Simon
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-06-07 15:19 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-06-03 14:57 [PATCH] [gdb.base] Convert multi-line function call into in foll-exec.c Kumar N, Bhuvanendra
2021-06-03 15:38 ` Simon Marchi
2021-06-07 6:31 ` Kumar N, Bhuvanendra
2021-06-07 15:07 ` Simon Marchi
2021-06-07 15:19 ` Kumar N, Bhuvanendra
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 read-only IMAP folder(s) and NNTP newsgroup(s).