public inbox for systemtap@sourceware.org
 help / color / mirror / Atom feed
From: "houtao1 at huawei dot com" <sourceware-bugzilla@sourceware.org>
To: systemtap@sourceware.org
Subject: [Bug translator/23775] Pass 2 failed when generate a kernel module for stap script which uses kernel trace-point
Date: Sun, 14 Oct 2018 11:29:00 -0000	[thread overview]
Message-ID: <bug-23775-6586-TjHRaeRYZb@http.sourceware.org/bugzilla/> (raw)
In-Reply-To: <bug-23775-6586@http.sourceware.org/bugzilla/>

https://sourceware.org/bugzilla/show_bug.cgi?id=23775

--- Comment #1 from houtao1 at huawei dot com ---
>Ignoring duplicate kernel source tree (DW_AT_comp_dir) at '/home/htbegin/code/linux/newest'
>Checking tracepoint glob /home/htbegin/code/linux/newest//include/trace/events/*.h
>Checking tracepoint glob /home/htbegin/code/linux/newest//include/trace/*.h
>Checking tracepoint glob /home/htbegin/code/linux/newest//include/ras/*_event.h
It seems that elaborator can not find the correct kernel source directory, so
it can not find the definition of trace-point block_rq_issue neither.

The problem can be fixed by the patch proposed in
https://sourceware.org/ml/systemtap/2018-q3/msg00166.html:

When generating kernel module for a systemtap script that uses trace-point
probe, if the vanilla kernel is built by using O=build_path option and
r=build_path option is passed to stap, stap will not be able to find
kernel_source_tree and will fail on pass-2.

Linux kernel will create a symlink named source to the source tree
for out-of-source build since (399b835be30e "kbuild: add a symlink
to the source for separate objdirs"), so fix the problem by checking
whether or not the symlink exists and using it as the kernel_source_tree.

Also using a new helper dir_exists() instead of file_exists() to
ensure the existence of the directory of source tree.

Signed-off-by: Hou Tao <houtao1@huawei.com>
---
 tapsets.cxx | 34 ++++++++++++++++++++++++++--------
 util.cxx    | 13 +++++++++++++
 util.h      |  1 +
 3 files changed, 40 insertions(+), 8 deletions(-)

diff --git a/tapsets.cxx b/tapsets.cxx
index ca94be4ce..8617796d8 100644
--- a/tapsets.cxx
+++ b/tapsets.cxx
@@ -12391,16 +12391,34 @@ tracepoint_builder::init_dw(systemtap_session& s)
     }

   // find kernel_source_tree from a source link, when different from build
-  if (s.kernel_source_tree == "" && endswith(s.kernel_build_tree, "/build"))
+  if (s.kernel_source_tree == "")
     {
-      string source_tree = s.kernel_build_tree;
-      source_tree.replace(source_tree.length() - 5, 5, "source");
-      if (file_exists(source_tree) &&
-          resolve_path(source_tree) != resolve_path(s.kernel_build_tree))
+      vector<string> source_trees;
+
+      // vendor kernel (e.g. Fedora): the source link is in the same dir
+      // as the build tree
+      if (endswith(s.kernel_build_tree, "/build"))
+        {
+          string source_tree = s.kernel_build_tree;
+          source_tree.replace(source_tree.length() - 5, 5, "source");
+          source_trees.push_back(source_tree);
+        }
+
+      // vanilla kernel: the source link is in the build tree
+      source_trees.push_back(s.kernel_build_tree + "/source");
+
+      for (unsigned i = 0; i < source_trees.size(); i++)
         {
-          if (s.verbose > 2)
-            clog << _F("Located kernel source tree at '%s'",
source_tree.c_str()) << endl;
-          s.kernel_source_tree = source_tree;
+          string source_tree = source_trees[i];
+
+          if (dir_exists(source_tree) &&
+              resolve_path(source_tree) != resolve_path(s.kernel_build_tree))
+            {
+              if (s.verbose > 2)
+                clog << _F("Located kernel source tree at '%s'",
source_tree.c_str()) << endl;
+              s.kernel_source_tree = source_tree;
+              break;
+            }
         }
     }

diff --git a/util.cxx b/util.cxx
index 1b1127995..b4f1e9c25 100644
--- a/util.cxx
+++ b/util.cxx
@@ -103,6 +103,19 @@ file_exists (const string &path)
   return false;
 }

+// Check that a dir is present
+bool
+dir_exists(const string &path)
+{
+  struct stat info;
+
+  if (stat(path.c_str(), &info) == 0 &&
+      S_ISDIR(info.st_mode))
+    return true;
+
+  return false;
+}
+
 // Copy a file.  The copy is done via a temporary file and atomic
 // rename.
 bool
diff --git a/util.h b/util.h
index b4c4e40b7..4e30d4237 100644
--- a/util.h
+++ b/util.h
@@ -72,6 +72,7 @@ const char *get_home_directory(void);
 size_t get_file_size(const std::string &path);
 size_t get_file_size(int fd);
 bool file_exists (const std::string &path);
+bool dir_exists(const std::string &path);
 bool copy_file(const std::string& src, const std::string& dest,
                bool verbose=false);
 int create_dir(const char *dir, int mode = 0777);
-- 
2.16.2.dirty

-- 
You are receiving this mail because:
You are the assignee for the bug.

  reply	other threads:[~2018-10-14 11:29 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-10-14 11:13 [Bug translator/23775] New: " houtao1 at huawei dot com
2018-10-14 11:29 ` houtao1 at huawei dot com [this message]
2018-10-14 11:41 ` [Bug translator/23775] " houtao1 at huawei dot com
2018-10-16 15:25 ` fche at redhat dot com

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bug-23775-6586-TjHRaeRYZb@http.sourceware.org/bugzilla/ \
    --to=sourceware-bugzilla@sourceware.org \
    --cc=systemtap@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).