[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 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