[PATCH] test: basename filenames in dwarf2 frame checks

[PATCH] test: basename filenames in dwarf2 frame checks

[Bratislav Filipovic]

Some gdb.dwarf2 tests match the output of commands like frame and
expect to see locations printed as fission-base.c:LINE (no directory
prefix).  When the test programs are built with clang as CC_FOR_TARGET,
GDB can print an absolute source path instead, causing FAILs.

For example, with clang:

    (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
    (gdb) frame
    #0  main () at /path/to/gdb/testsuite/fission-base.c:27
    27        return func (-1);
    (gdb) FAIL: gdb.dwarf2/fission-base.exp: frame in main

With gcc:

    (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
    (gdb) frame
    #0  main () at fission-base.c:27
    27        return func (-1);
    (gdb) PASS: gdb.dwarf2/fission-base.exp: frame in main

The difference comes from the DWARF line table contents.  With clang,
the .debug_line directory table can contain an absolute directory that
the file table entries reference:

    The Directory Table:
      0  /path/to/gdb/testsuite

    The File Name Table:
      0  Dir=0  Name=fission-base.c
      1  Dir=0  Name=fission-base.c

Whereas with gcc the directory table is empty and only the bare filename
is present:

    The Directory Table is empty.

    The File Name Table:
      1  Name=fission-base.c

This difference reflects toolchain/assembler behavior in how .file/.loc
are translated into .debug_line and is orthogonal to what these tests
aim to validate.

Force set filename-display basename in the affected tests so the
output is stable across toolchains.

Test: gdb.dwarf2/fission-base.exp (CC_FOR_TARGET=clang, gcc)
Test: gdb.dwarf2/dw2-undefined-ret-addr.exp (CC_FOR_TARGET=clang, gcc)
---
 gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp | 5 +++++
 gdb/testsuite/gdb.dwarf2/fission-base.exp           | 5 +++++
 2 files changed, 10 insertions(+)

diff --git a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
index 32ddaf4a..5b47c84c 100644
--- a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
+++ b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
@@ -30,6 +30,11 @@ if {![runto "stop_frame"]} {
     return -1
 }
 
+# If test is compiled with clang, GDB would display absolute path
+# This command keeps output consistent across toolchains.
+
+gdb_test_no_output "set filename-display basename"
+
 # stop_frame should be the outermost frame.
 
 # Check that backtrace shows only frame #0.
diff --git a/gdb/testsuite/gdb.dwarf2/fission-base.exp b/gdb/testsuite/gdb.dwarf2/fission-base.exp
index 035fe731..a73cab48 100644
--- a/gdb/testsuite/gdb.dwarf2/fission-base.exp
+++ b/gdb/testsuite/gdb.dwarf2/fission-base.exp
@@ -46,6 +46,11 @@ if {![runto_main]} {
 gdb_test "ptype main" "type = int \\(\\)"
 gdb_test "ptype func" "type = int \\(int\\)"
 
+# If test is compiled with clang, GDB would display absolute path
+# This command keeps output consistent across toolchains.
+
+gdb_test_no_output "set filename-display basename"
+
 gdb_test "frame" "#0 *main \\(\\) at $testfile\\.c:$decimal.*" \
     "frame in main"
 
-- 
2.43.0

RE: [PING][PATCH] test: basename filenames in dwarf2 frame checks

[Filipovic, Bratislav]

[AMD Official Use Only - AMD Internal Distribution Only]

PING
Can someone take a look at this?

Regards Bratislav

-----Original Message-----
From: Filipovic, Bratislav <Bratislav.Filipovic@amd.com>
Sent: Wednesday, March 25, 2026 5:32 PM
To: gdb-patches@sourceware.org
Cc: simon.marchi@efficios.com; Filipovic, Bratislav <Bratislav.Filipovic@amd.com>
Subject: [PATCH] test: basename filenames in dwarf2 frame checks

Some gdb.dwarf2 tests match the output of commands like frame and expect to see locations printed as fission-base.c:LINE (no directory prefix).  When the test programs are built with clang as CC_FOR_TARGET, GDB can print an absolute source path instead, causing FAILs.

For example, with clang:

    (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
    (gdb) frame
    #0  main () at /path/to/gdb/testsuite/fission-base.c:27
    27        return func (-1);
    (gdb) FAIL: gdb.dwarf2/fission-base.exp: frame in main

With gcc:

    (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
    (gdb) frame
    #0  main () at fission-base.c:27
    27        return func (-1);
    (gdb) PASS: gdb.dwarf2/fission-base.exp: frame in main

The difference comes from the DWARF line table contents.  With clang, the .debug_line directory table can contain an absolute directory that the file table entries reference:

    The Directory Table:
      0  /path/to/gdb/testsuite

    The File Name Table:
      0  Dir=0  Name=fission-base.c
      1  Dir=0  Name=fission-base.c

Whereas with gcc the directory table is empty and only the bare filename is present:

    The Directory Table is empty.

    The File Name Table:
      1  Name=fission-base.c

This difference reflects toolchain/assembler behavior in how .file/.loc are translated into .debug_line and is orthogonal to what these tests aim to validate.

Force set filename-display basename in the affected tests so the output is stable across toolchains.

Test: gdb.dwarf2/fission-base.exp (CC_FOR_TARGET=clang, gcc)
Test: gdb.dwarf2/dw2-undefined-ret-addr.exp (CC_FOR_TARGET=clang, gcc)
---
 gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp | 5 +++++
 gdb/testsuite/gdb.dwarf2/fission-base.exp           | 5 +++++
 2 files changed, 10 insertions(+)

diff --git a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
index 32ddaf4a..5b47c84c 100644
--- a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
+++ b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
@@ -30,6 +30,11 @@ if {![runto "stop_frame"]} {
     return -1
 }

+# If test is compiled with clang, GDB would display absolute path #
+This command keeps output consistent across toolchains.
+
+gdb_test_no_output "set filename-display basename"
+
 # stop_frame should be the outermost frame.

 # Check that backtrace shows only frame #0.
diff --git a/gdb/testsuite/gdb.dwarf2/fission-base.exp b/gdb/testsuite/gdb.dwarf2/fission-base.exp
index 035fe731..a73cab48 100644
--- a/gdb/testsuite/gdb.dwarf2/fission-base.exp
+++ b/gdb/testsuite/gdb.dwarf2/fission-base.exp
@@ -46,6 +46,11 @@ if {![runto_main]} {
 gdb_test "ptype main" "type = int \\(\\)"
 gdb_test "ptype func" "type = int \\(int\\)"

+# If test is compiled with clang, GDB would display absolute path #
+This command keeps output consistent across toolchains.
+
+gdb_test_no_output "set filename-display basename"
+
 gdb_test "frame" "#0 *main \\(\\) at $testfile\\.c:$decimal.*" \
     "frame in main"

--
2.43.0

Re: [PATCH] test: basename filenames in dwarf2 frame checks

[Andrew Burgess]

Bratislav Filipovic <bfilipov@amd.com> writes:

> Some gdb.dwarf2 tests match the output of commands like frame and
> expect to see locations printed as fission-base.c:LINE (no directory
> prefix).  When the test programs are built with clang as CC_FOR_TARGET,
> GDB can print an absolute source path instead, causing FAILs.

Thanks for taking a look at this, it's always nice to see tests being
improved like this.

This is fine, with two minor nits fixed...

>
> diff --git a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> index 32ddaf4a..5b47c84c 100644
> --- a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> +++ b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> @@ -30,6 +30,11 @@ if {![runto "stop_frame"]} {
>      return -1
>  }
>  
> +# If test is compiled with clang, GDB would display absolute path
> +# This command keeps output consistent across toolchains.

This comment needs to be one complete sentence (add a comma after
'path', and change 'This' to 'this'), or two sentences (add a full stop
after 'path').

> +
> +gdb_test_no_output "set filename-display basename"
> +
>  # stop_frame should be the outermost frame.
>  
>  # Check that backtrace shows only frame #0.
> diff --git a/gdb/testsuite/gdb.dwarf2/fission-base.exp b/gdb/testsuite/gdb.dwarf2/fission-base.exp
> index 035fe731..a73cab48 100644
> --- a/gdb/testsuite/gdb.dwarf2/fission-base.exp
> +++ b/gdb/testsuite/gdb.dwarf2/fission-base.exp
> @@ -46,6 +46,11 @@ if {![runto_main]} {
>  gdb_test "ptype main" "type = int \\(\\)"
>  gdb_test "ptype func" "type = int \\(int\\)"
>  
> +# If test is compiled with clang, GDB would display absolute path
> +# This command keeps output consistent across toolchains.

Same issue here.

With both of those fixed:

Approved-By: Andrew Burgess <aburgess@redhat.com>

thanks,
Andrew

[PATCH] test: basename filenames in dwarf2 frame checks

[Bratislav Filipovic]

Some gdb.dwarf2 tests match the output of commands like frame and
expect to see locations printed as fission-base.c:LINE (no directory
prefix).  When the test programs are built with clang as CC_FOR_TARGET,
GDB can print an absolute source path instead, causing FAILs.

For example, with clang:

    (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
    (gdb) frame
    #0  main () at /path/to/gdb/testsuite/fission-base.c:27
    27        return func (-1);
    (gdb) FAIL: gdb.dwarf2/fission-base.exp: frame in main

With gcc:

    (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
    (gdb) frame
    #0  main () at fission-base.c:27
    27        return func (-1);
    (gdb) PASS: gdb.dwarf2/fission-base.exp: frame in main

The difference comes from the DWARF line table contents.  With clang,
the .debug_line directory table can contain an absolute directory that
the file table entries reference:

    The Directory Table:
      0  /path/to/gdb/testsuite

    The File Name Table:
      0  Dir=0  Name=fission-base.c
      1  Dir=0  Name=fission-base.c

Whereas with gcc the directory table is empty and only the bare filename
is present:

    The Directory Table is empty.

    The File Name Table:
      1  Name=fission-base.c

This difference reflects toolchain/assembler behavior in how .file/.loc
are translated into .debug_line and is orthogonal to what these tests
aim to validate.

Force set filename-display basename in the affected tests so the
output is stable across toolchains.

Test: gdb.dwarf2/fission-base.exp (CC_FOR_TARGET=clang, gcc)
Test: gdb.dwarf2/dw2-undefined-ret-addr.exp (CC_FOR_TARGET=clang, gcc)
---

Thanks for the review Andrew. Fixed the comments.

Regards Bratislav


 gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp | 5 +++++
 gdb/testsuite/gdb.dwarf2/fission-base.exp           | 5 +++++
 2 files changed, 10 insertions(+)

diff --git a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
index 32ddaf4a..88765d4f 100644
--- a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
+++ b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
@@ -30,6 +30,11 @@ if {![runto "stop_frame"]} {
     return -1
 }
 
+# If test is compiled with clang, GDB would display absolute path.
+# This command keeps output consistent across toolchains.
+
+gdb_test_no_output "set filename-display basename"
+
 # stop_frame should be the outermost frame.
 
 # Check that backtrace shows only frame #0.
diff --git a/gdb/testsuite/gdb.dwarf2/fission-base.exp b/gdb/testsuite/gdb.dwarf2/fission-base.exp
index 035fe731..d1abf6c9 100644
--- a/gdb/testsuite/gdb.dwarf2/fission-base.exp
+++ b/gdb/testsuite/gdb.dwarf2/fission-base.exp
@@ -46,6 +46,11 @@ if {![runto_main]} {
 gdb_test "ptype main" "type = int \\(\\)"
 gdb_test "ptype func" "type = int \\(int\\)"
 
+# If test is compiled with clang, GDB would display absolute path.
+# This command keeps output consistent across toolchains.
+
+gdb_test_no_output "set filename-display basename"
+
 gdb_test "frame" "#0 *main \\(\\) at $testfile\\.c:$decimal.*" \
     "frame in main"
 
-- 
2.43.0

Re: [PATCH] test: basename filenames in dwarf2 frame checks

[Andrew Burgess]

Bratislav Filipovic <bfilipov@amd.com> writes:

> Some gdb.dwarf2 tests match the output of commands like frame and
> expect to see locations printed as fission-base.c:LINE (no directory
> prefix).  When the test programs are built with clang as CC_FOR_TARGET,
> GDB can print an absolute source path instead, causing FAILs.
>
> For example, with clang:
>
>     (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
>     (gdb) frame
>     #0  main () at /path/to/gdb/testsuite/fission-base.c:27
>     27        return func (-1);
>     (gdb) FAIL: gdb.dwarf2/fission-base.exp: frame in main
>
> With gcc:
>
>     (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
>     (gdb) frame
>     #0  main () at fission-base.c:27
>     27        return func (-1);
>     (gdb) PASS: gdb.dwarf2/fission-base.exp: frame in main
>
> The difference comes from the DWARF line table contents.  With clang,
> the .debug_line directory table can contain an absolute directory that
> the file table entries reference:
>
>     The Directory Table:
>       0  /path/to/gdb/testsuite
>
>     The File Name Table:
>       0  Dir=0  Name=fission-base.c
>       1  Dir=0  Name=fission-base.c
>
> Whereas with gcc the directory table is empty and only the bare filename
> is present:
>
>     The Directory Table is empty.
>
>     The File Name Table:
>       1  Name=fission-base.c
>
> This difference reflects toolchain/assembler behavior in how .file/.loc
> are translated into .debug_line and is orthogonal to what these tests
> aim to validate.
>
> Force set filename-display basename in the affected tests so the
> output is stable across toolchains.
>
> Test: gdb.dwarf2/fission-base.exp (CC_FOR_TARGET=clang, gcc)
> Test: gdb.dwarf2/dw2-undefined-ret-addr.exp (CC_FOR_TARGET=clang, gcc)
> ---
>
> Thanks for the review Andrew. Fixed the comments.

Looks good.  Feel free to push this patch.

Approved-By: Andrew Burgess <aburgess@redhat.com>

Thanks,
Andrew


>
> Regards Bratislav
>
>
>  gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp | 5 +++++
>  gdb/testsuite/gdb.dwarf2/fission-base.exp           | 5 +++++
>  2 files changed, 10 insertions(+)
>
> diff --git a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> index 32ddaf4a..88765d4f 100644
> --- a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> +++ b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> @@ -30,6 +30,11 @@ if {![runto "stop_frame"]} {
>      return -1
>  }
>  
> +# If test is compiled with clang, GDB would display absolute path.
> +# This command keeps output consistent across toolchains.
> +
> +gdb_test_no_output "set filename-display basename"
> +
>  # stop_frame should be the outermost frame.
>  
>  # Check that backtrace shows only frame #0.
> diff --git a/gdb/testsuite/gdb.dwarf2/fission-base.exp b/gdb/testsuite/gdb.dwarf2/fission-base.exp
> index 035fe731..d1abf6c9 100644
> --- a/gdb/testsuite/gdb.dwarf2/fission-base.exp
> +++ b/gdb/testsuite/gdb.dwarf2/fission-base.exp
> @@ -46,6 +46,11 @@ if {![runto_main]} {
>  gdb_test "ptype main" "type = int \\(\\)"
>  gdb_test "ptype func" "type = int \\(int\\)"
>  
> +# If test is compiled with clang, GDB would display absolute path.
> +# This command keeps output consistent across toolchains.
> +
> +gdb_test_no_output "set filename-display basename"
> +
>  gdb_test "frame" "#0 *main \\(\\) at $testfile\\.c:$decimal.*" \
>      "frame in main"
>  
> -- 
> 2.43.0

RE: [PATCH] test: basename filenames in dwarf2 frame checks

[Filipovic, Bratislav]

[AMD Official Use Only - AMD Internal Distribution Only]

Hello Andrew,

I still don't have write-after-approval yet. Can you get it pushed?

Regards Bratislav

-----Original Message-----
From: Andrew Burgess <aburgess@redhat.com>
Sent: Thursday, April 2, 2026 2:53 PM
To: Filipovic, Bratislav <Bratislav.Filipovic@amd.com>
Cc: Filipovic, Bratislav <Bratislav.Filipovic@amd.com>; gdb-patches@sourceware.org; simon.marchi@efficios.com
Subject: Re: [PATCH] test: basename filenames in dwarf2 frame checks

[You don't often get email from aburgess@redhat.com. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]

Caution: This message originated from an External Source. Use proper caution when opening attachments, clicking links, or responding.


Bratislav Filipovic <bfilipov@amd.com> writes:

> Some gdb.dwarf2 tests match the output of commands like frame and
> expect to see locations printed as fission-base.c:LINE (no directory
> prefix).  When the test programs are built with clang as
> CC_FOR_TARGET, GDB can print an absolute source path instead, causing FAILs.
>
> For example, with clang:
>
>     (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
>     (gdb) frame
>     #0  main () at /path/to/gdb/testsuite/fission-base.c:27
>     27        return func (-1);
>     (gdb) FAIL: gdb.dwarf2/fission-base.exp: frame in main
>
> With gcc:
>
>     (gdb) PASS: gdb.dwarf2/fission-base.exp: ptype func
>     (gdb) frame
>     #0  main () at fission-base.c:27
>     27        return func (-1);
>     (gdb) PASS: gdb.dwarf2/fission-base.exp: frame in main
>
> The difference comes from the DWARF line table contents.  With clang,
> the .debug_line directory table can contain an absolute directory that
> the file table entries reference:
>
>     The Directory Table:
>       0  /path/to/gdb/testsuite
>
>     The File Name Table:
>       0  Dir=0  Name=fission-base.c
>       1  Dir=0  Name=fission-base.c
>
> Whereas with gcc the directory table is empty and only the bare
> filename is present:
>
>     The Directory Table is empty.
>
>     The File Name Table:
>       1  Name=fission-base.c
>
> This difference reflects toolchain/assembler behavior in how
> .file/.loc are translated into .debug_line and is orthogonal to what
> these tests aim to validate.
>
> Force set filename-display basename in the affected tests so the
> output is stable across toolchains.
>
> Test: gdb.dwarf2/fission-base.exp (CC_FOR_TARGET=clang, gcc)
> Test: gdb.dwarf2/dw2-undefined-ret-addr.exp (CC_FOR_TARGET=clang, gcc)
> ---
>
> Thanks for the review Andrew. Fixed the comments.

Looks good.  Feel free to push this patch.

Approved-By: Andrew Burgess <aburgess@redhat.com>

Thanks,
Andrew


>
> Regards Bratislav
>
>
>  gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp | 5 +++++
>  gdb/testsuite/gdb.dwarf2/fission-base.exp           | 5 +++++
>  2 files changed, 10 insertions(+)
>
> diff --git a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> index 32ddaf4a..88765d4f 100644
> --- a/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> +++ b/gdb/testsuite/gdb.dwarf2/dw2-undefined-ret-addr.exp
> @@ -30,6 +30,11 @@ if {![runto "stop_frame"]} {
>      return -1
>  }
>
> +# If test is compiled with clang, GDB would display absolute path.
> +# This command keeps output consistent across toolchains.
> +
> +gdb_test_no_output "set filename-display basename"
> +
>  # stop_frame should be the outermost frame.
>
>  # Check that backtrace shows only frame #0.
> diff --git a/gdb/testsuite/gdb.dwarf2/fission-base.exp
> b/gdb/testsuite/gdb.dwarf2/fission-base.exp
> index 035fe731..d1abf6c9 100644
> --- a/gdb/testsuite/gdb.dwarf2/fission-base.exp
> +++ b/gdb/testsuite/gdb.dwarf2/fission-base.exp
> @@ -46,6 +46,11 @@ if {![runto_main]} {  gdb_test "ptype main" "type =
> int \\(\\)"
>  gdb_test "ptype func" "type = int \\(int\\)"
>
> +# If test is compiled with clang, GDB would display absolute path.
> +# This command keeps output consistent across toolchains.
> +
> +gdb_test_no_output "set filename-display basename"
> +
>  gdb_test "frame" "#0 *main \\(\\) at $testfile\\.c:$decimal.*" \
>      "frame in main"
>
> --
> 2.43.0