If BUILD_SRC or CURDIR contains tailing '/', the file names passed to gcc will
contain '//'. It will be contained .o's in debuginfo, then confuse debugedit:
https://bugzilla.redhat.com/show_bug.cgi?id=304121
This patch uses realpath command to makesure potential tailing '/'s are removed.
Signed-off-by: Wang Nan <[email protected]>
Cc: Arnaldo Carvalho de Melo <[email protected]>
Cc: Jiri Olsa <[email protected]>
Cc: Steven Rostedt <[email protected]>
Cc: Namhyung Kim <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: Geng Hui <[email protected]>
---
tools/lib/traceevent/Makefile | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/lib/traceevent/Makefile b/tools/lib/traceevent/Makefile
index 005c9cc..5bb281c 100644
--- a/tools/lib/traceevent/Makefile
+++ b/tools/lib/traceevent/Makefile
@@ -107,8 +107,8 @@ endif # BUILD_SRC
# We process the rest of the Makefile if this is the final invocation of make
ifeq ($(skip-makefile),)
-srctree := $(if $(BUILD_SRC),$(BUILD_SRC),$(CURDIR))
-objtree := $(CURDIR)
+srctree := $(realpath $(if $(BUILD_SRC),$(BUILD_SRC),$(CURDIR)))
+objtree := $(realpath $(CURDIR))
src := $(srctree)
obj := $(objtree)
--
1.8.4
Another email inbox clean up: BLAST FROM THE PAST
On Fri, 7 Feb 2014 12:22:33 +0800
Wang Nan <[email protected]> wrote:
> If BUILD_SRC or CURDIR contains tailing '/', the file names passed to gcc will
> contain '//'. It will be contained .o's in debuginfo, then confuse debugedit:
>
> https://bugzilla.redhat.com/show_bug.cgi?id=304121
>
> This patch uses realpath command to makesure potential tailing '/'s are removed.
Is this still an issue? I know Jiri did a bunch of changes to the make
process, but does this still confuse debugedit (whatever that is)?
-- Steve
>
> Signed-off-by: Wang Nan <[email protected]>
> Cc: Arnaldo Carvalho de Melo <[email protected]>
> Cc: Jiri Olsa <[email protected]>
> Cc: Steven Rostedt <[email protected]>
> Cc: Namhyung Kim <[email protected]>
> Cc: Ingo Molnar <[email protected]>
> Cc: Andrew Morton <[email protected]>
> Cc: Geng Hui <[email protected]>
>
> ---
> tools/lib/traceevent/Makefile | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/tools/lib/traceevent/Makefile b/tools/lib/traceevent/Makefile
> index 005c9cc..5bb281c 100644
> --- a/tools/lib/traceevent/Makefile
> +++ b/tools/lib/traceevent/Makefile
> @@ -107,8 +107,8 @@ endif # BUILD_SRC
> # We process the rest of the Makefile if this is the final invocation of make
> ifeq ($(skip-makefile),)
>
> -srctree := $(if $(BUILD_SRC),$(BUILD_SRC),$(CURDIR))
> -objtree := $(CURDIR)
> +srctree := $(realpath $(if $(BUILD_SRC),$(BUILD_SRC),$(CURDIR)))
> +objtree := $(realpath $(CURDIR))
> src := $(srctree)
> obj := $(objtree)
>
On Mon, Feb 23, 2015 at 09:09:32PM -0500, Steven Rostedt wrote:
> Another email inbox clean up: BLAST FROM THE PAST
>
> On Fri, 7 Feb 2014 12:22:33 +0800
> Wang Nan <[email protected]> wrote:
>
> > If BUILD_SRC or CURDIR contains tailing '/', the file names passed to gcc will
> > contain '//'. It will be contained .o's in debuginfo, then confuse debugedit:
> >
> > https://bugzilla.redhat.com/show_bug.cgi?id=304121
> >
> > This patch uses realpath command to makesure potential tailing '/'s are removed.
>
> Is this still an issue? I know Jiri did a bunch of changes to the make
> process, but does this still confuse debugedit (whatever that is)?
this should no longer be an issue.. recent build changes
take care of this
jirka