Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Pedro Alves <palves@redhat.com>
To: Yao Qi <yao@codesourcery.com>
Cc: gdb-patches@sourceware.org
Subject: [PATCH] Fix regular /path/to/directory sysroots and target reported dll paths with drive specs. (was: Re: [PATCH] Don't set dir separator if path has drive spec)
Date: Fri, 27 Sep 2013 12:58:00 -0000	[thread overview]
Message-ID: <524580D8.3090409@redhat.com> (raw)
In-Reply-To: <1370306007-877-1-git-send-email-yao@codesourcery.com>

On 06/04/2013 01:33 AM, Yao Qi wrote:
> Hi,
> When I debug a remote Windows program on Linux host, set sysroot to
> "remote:", find GDB can't load these dlls from the remote.  If the dll
> name is "C:/Windows/system32/ntdll.dll" and GDB add "remote:" prefix
> to the path.  The dll name becomes "remote:/C:/Windows/system32/ntdll.dll"
> and GDBserver is confused by the name.  The first slash shouldn't be
> added.  In solib.c:solib_find,
> 
>       need_dir_separator = !IS_DIR_SEPARATOR (in_pathname[0]);
> 
> it controls whether to insert dir separator after sysroot.  It works
> for UNIX-like file path, but it doesn't work for the DOS-like absolute
> path.  This patch is to fix this problem.  Is it OK?
> 
> gdb:
> 
> 2013-06-04  Yao Qi  <yao@codesourcery.com>
> 
> 	* solib.c (solib_find): Don't need dir separator if path has
> 	drive spec.

This actually broke sysroots that are _not_ "remote:".

Comments on the patch below?

------------
Subject: [PATCH] Fix regular /path/to/directory sysroots and target reported
 dll paths with drive specs.

I tried debugging a remote Windows program on Linux host, and pointed the
sysroot to "/some/path/" rather than "remote:", and I found GDB couldn't
find the dlls in the sysroot.  If the dll name is
"C:/Windows/system32/ntdll.dll", I end up with the sysroot+in_pathname
concatenated this way:

 (top-gdb) p temp_pathname
 $1 = 0x228b690 "/some/pathC:/Windows/system32/ntdll.dll"
                          ^^

That is, a directory separator is missing.  This is a regression.

The problem is that solib_find decides that since the target path has
a drive spec, a separator is not necessary, which is clearly wrong in
this case.  That check was added in
<https://sourceware.org/ml/gdb-patches/2013-06/msg00028.html>, to
handle the case of sysroot being "remote:".  This patch fixes that
original issue in a different way.  Instead of checking whether the
path has a drive spec, check whether the sysroot is "remote:".  The
patch adds a table that helps visualize the cases that need a
separator.  I also confirmed the original issue is still handled as
expected.  That is, that "set sysroot remote:" still does the right
thing.

remote_filename_p returns true if the filename is prefixed with
"remote:".  In this case, we need to check whether the filename is
exactly "remote:".  I thought of different ways or either changing
remote_filename_p or adding another convenience function to remote.c
to avoid exposing the "remote:" prefix out of remote.c.  But all
attempts turned out adding lot of over needless complication.  So the
patch just exposes the prefix behind a new macro, which allows using a
straighforward strcmp.

gdb/
2013-09-27  Pedro Alves  <palves@redhat.com>

	* remote.h (REMOTE_SYSROOT_PREFIX): New define.
	(remote_filename_p): Add comment.
	* remote.c (remote_filename_p): Adjust to use
	REMOTE_SYSROOT_PREFIX.
	* solib.c (solib_find): When deciding whether we need to add a
	directory separator, check whether the sysroot is "remote:"
	instead of checking whether the patch has a drive spec.  Add
	comments.
---
 gdb/remote.c |  4 +++-
 gdb/remote.h |  7 +++++++
 gdb/solib.c  | 21 +++++++++++++++++++--
 3 files changed, 29 insertions(+), 3 deletions(-)

diff --git a/gdb/remote.c b/gdb/remote.c
index 0fa1e2b..a9ef297 100644
--- a/gdb/remote.c
+++ b/gdb/remote.c
@@ -10036,7 +10036,9 @@ remote_bfd_iovec_stat (struct bfd *abfd, void *stream, struct stat *sb)
 int
 remote_filename_p (const char *filename)
 {
-  return strncmp (filename, "remote:", 7) == 0;
+  return strncmp (filename,
+		  REMOTE_SYSROOT_PREFIX,
+		  sizeof (REMOTE_SYSROOT_PREFIX) - 1) == 0;
 }
 
 bfd *
diff --git a/gdb/remote.h b/gdb/remote.h
index c036db7..b6ea547 100644
--- a/gdb/remote.h
+++ b/gdb/remote.h
@@ -57,6 +57,13 @@ void remote_file_delete (const char *remote_file, int from_tty);
 
 bfd *remote_bfd_open (const char *remote_file, const char *target);
 
+/* If a path starts with this sequence, GDB will retrieve the target
+   libraries from the remote system.  */
+
+#define REMOTE_SYSROOT_PREFIX "remote:"
+
+/* True if FILENAME starts with REMOTE_SYSROOT_PREFIX.  */
+
 int remote_filename_p (const char *filename);
 
 extern int remote_register_number_and_offset (struct gdbarch *gdbarch,
diff --git a/gdb/solib.c b/gdb/solib.c
index 5456d6f..5516b3e 100644
--- a/gdb/solib.c
+++ b/gdb/solib.c
@@ -230,8 +230,25 @@ solib_find (char *in_pathname, int *fd)
     {
       int need_dir_separator;
 
-      need_dir_separator = (!IS_DIR_SEPARATOR (in_pathname[0])
-			    && !HAS_TARGET_DRIVE_SPEC (fskind, in_pathname));
+      /* Concatenate the sysroot and the target reported filename.  We
+	 may need to glue them with a directory separator.  Cases to
+	 consider:
+
+        | sysroot         | separator | in_pathname    |
+        |-----------------+-----------+----------------|
+        | /some/dir       | /         | c:/foo/bar.dll |
+        | /some/dir       |           | /foo/bar.dll   |
+        | remote:         |           | c:/foo/bar.dll |
+        | remote:         |           | /foo/bar.dll   |
+        | remote:some/dir | /         | c:/foo/bar.dll |
+        | remote:some/dir |           | /foo/bar.dll   |
+
+	IOW, we don't need to add a separator if IN_PATHNAME already
+	has one, or when the the sysroot is exactly "remote:".
+	There's no need to check for drive spec explicitly, as we only
+	get here if IN_PATHNAME is considered an absolute path.  */
+      need_dir_separator = !(IS_DIR_SEPARATOR (in_pathname[0])
+			     || strcmp (REMOTE_SYSROOT_PREFIX, sysroot) == 0);
 
       /* Cat the prefixed pathname together.  */
       temp_pathname = concat (sysroot,
-- 
1.7.11.7



  parent reply	other threads:[~2013-09-27 12:58 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-06-04  0:32 [PATCH] Don't set dir separator if path has drive spec Yao Qi
2013-06-04  0:48 ` Doug Evans
2013-06-04  1:33   ` Yao Qi
2013-09-27 12:58 ` Pedro Alves [this message]
2013-09-27 14:18   ` [PATCH] Fix regular /path/to/directory sysroots and target reported dll paths with drive specs Yao Qi
2013-09-27 15:29     ` Pedro Alves

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=524580D8.3090409@redhat.com \
    --to=palves@redhat.com \
    --cc=gdb-patches@sourceware.org \
    --cc=yao@codesourcery.com \
    /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