* [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).