From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Cooper Subject: Re: [PATCH] x86: print relevant (tail) part of filename for warnings and crashes Date: Wed, 16 Oct 2013 16:57:23 +0100 Message-ID: <525EB763.6030104@citrix.com> References: <525ECC1702000078000FB8D7@nat28.tlf.novell.com> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============8862587905135692994==" Return-path: Received: from mail6.bemta3.messagelabs.com ([195.245.230.39]) by lists.xen.org with esmtp (Exim 4.72) (envelope-from ) id 1VWTTU-0001Im-63 for xen-devel@lists.xenproject.org; Wed, 16 Oct 2013 15:57:28 +0000 In-Reply-To: <525ECC1702000078000FB8D7@nat28.tlf.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Jan Beulich Cc: xen-devel , Keir Fraser List-Id: xen-devel@lists.xenproject.org --===============8862587905135692994== Content-Type: multipart/alternative; boundary="------------030204050404030606080807" --------------030204050404030606080807 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit On 16/10/13 16:25, Jan Beulich wrote: > In particular when the origin construct is in a header file (and > hence the file name is an absolute path instead of just the file name > portion) the information can otherwise become rather useless when the > build tree isn't sitting relatively close to the file system root. > > Signed-off-by: Jan Beulich I had encountered similar frustrations in the past and had looked at a different solution to the problem. At the moment, __FILE__ is absolute, (and redundant many times in the image, although I guess it should compress well). I had investigated running a makefile rune to strip XEN_ROOT off $(abspath $<) and provide this as a __FILENAME__ macro, but I wasn't certain I had done it correctly. This would cause a rather smaller data section and substantially less printed in each of these cases. The only problem would then come if we end up with more than 50 characters below XEN_ROOT, which is substantially less likely ~Andrew > > --- a/xen/arch/x86/traps.c > +++ b/xen/arch/x86/traps.c > @@ -945,7 +945,7 @@ void do_invalid_op(struct cpu_user_regs > { > const struct bug_frame *bug; > u8 bug_insn[2]; > - const char *filename, *predicate, *eip = (char *)regs->eip; > + const char *prefix = "", *filename, *predicate, *eip = (char *)regs->eip; > unsigned long fixup; > int id, lineno; > static const struct bug_frame *const stop_frames[] = { > @@ -996,21 +996,27 @@ void do_invalid_op(struct cpu_user_regs > filename = bug_ptr(bug); > if ( !is_kernel(filename) ) > goto die; > + fixup = strlen(filename); > + if ( fixup > 50 ) > + { > + filename += fixup - 47; > + prefix = "..."; > + } > lineno = bug_line(bug); > > switch ( id ) > { > case BUGFRAME_warn: > - printk("Xen WARN at %.50s:%d\n", filename, lineno); > + printk("Xen WARN at %s%s:%d\n", prefix, filename, lineno); > show_execution_state(regs); > regs->eip = (unsigned long)eip; > return; > > case BUGFRAME_bug: > - printk("Xen BUG at %.50s:%d\n", filename, lineno); > + printk("Xen BUG at %s%s:%d\n", prefix, filename, lineno); > DEBUGGER_trap_fatal(TRAP_invalid_op, regs); > show_execution_state(regs); > - panic("Xen BUG at %.50s:%d\n", filename, lineno); > + panic("Xen BUG at %s%s:%d\n", prefix, filename, lineno); > > case BUGFRAME_assert: > /* ASSERT: decode the predicate string pointer. */ > @@ -1018,12 +1024,12 @@ void do_invalid_op(struct cpu_user_regs > if ( !is_kernel(predicate) ) > predicate = ""; > > - printk("Assertion '%s' failed at %.50s:%d\n", > - predicate, filename, lineno); > + printk("Assertion '%s' failed at %s%s:%d\n", > + predicate, prefix, filename, lineno); > DEBUGGER_trap_fatal(TRAP_invalid_op, regs); > show_execution_state(regs); > - panic("Assertion '%s' failed at %.50s:%d\n", > - predicate, filename, lineno); > + panic("Assertion '%s' failed at %s%s:%d\n", > + predicate, prefix, filename, lineno); > } > > die: > > > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@lists.xen.org > http://lists.xen.org/xen-devel --------------030204050404030606080807 Content-Type: text/html; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit
On 16/10/13 16:25, Jan Beulich wrote:
In particular when the origin construct is in a header file (and
hence the file name is an absolute path instead of just the file name
portion) the information can otherwise become rather useless when the
build tree isn't sitting relatively close to the file system root.

Signed-off-by: Jan Beulich <jbeulich@suse.com>

I had encountered similar frustrations in the past and had looked at a different solution to the problem.

At the moment, __FILE__ is absolute, (and redundant many times in the image, although I guess it should compress well).

I had investigated running a makefile rune to strip XEN_ROOT off $(abspath $<) and provide this as a __FILENAME__ macro, but I wasn't certain I had done it correctly.

This would cause a rather smaller data section and substantially less printed in each of these cases.  The only problem would then come if we end up with more than 50 characters below XEN_ROOT, which is substantially less likely

~Andrew


--- a/xen/arch/x86/traps.c
+++ b/xen/arch/x86/traps.c
@@ -945,7 +945,7 @@ void do_invalid_op(struct cpu_user_regs 
 {
     const struct bug_frame *bug;
     u8 bug_insn[2];
-    const char *filename, *predicate, *eip = (char *)regs->eip;
+    const char *prefix = "", *filename, *predicate, *eip = (char *)regs->eip;
     unsigned long fixup;
     int id, lineno;
     static const struct bug_frame *const stop_frames[] = {
@@ -996,21 +996,27 @@ void do_invalid_op(struct cpu_user_regs 
     filename = bug_ptr(bug);
     if ( !is_kernel(filename) )
         goto die;
+    fixup = strlen(filename);
+    if ( fixup > 50 )
+    {
+        filename += fixup - 47;
+        prefix = "...";
+    }
     lineno = bug_line(bug);
 
     switch ( id )
     {
     case BUGFRAME_warn:
-        printk("Xen WARN at %.50s:%d\n", filename, lineno);
+        printk("Xen WARN at %s%s:%d\n", prefix, filename, lineno);
         show_execution_state(regs);
         regs->eip = (unsigned long)eip;
         return;
 
     case BUGFRAME_bug:
-        printk("Xen BUG at %.50s:%d\n", filename, lineno);
+        printk("Xen BUG at %s%s:%d\n", prefix, filename, lineno);
         DEBUGGER_trap_fatal(TRAP_invalid_op, regs);
         show_execution_state(regs);
-        panic("Xen BUG at %.50s:%d\n", filename, lineno);
+        panic("Xen BUG at %s%s:%d\n", prefix, filename, lineno);
 
     case BUGFRAME_assert:
         /* ASSERT: decode the predicate string pointer. */
@@ -1018,12 +1024,12 @@ void do_invalid_op(struct cpu_user_regs 
         if ( !is_kernel(predicate) )
             predicate = "<unknown>";
 
-        printk("Assertion '%s' failed at %.50s:%d\n",
-               predicate, filename, lineno);
+        printk("Assertion '%s' failed at %s%s:%d\n",
+               predicate, prefix, filename, lineno);
         DEBUGGER_trap_fatal(TRAP_invalid_op, regs);
         show_execution_state(regs);
-        panic("Assertion '%s' failed at %.50s:%d\n",
-              predicate, filename, lineno);
+        panic("Assertion '%s' failed at %s%s:%d\n",
+              predicate, prefix, filename, lineno);
     }
 
  die:






_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

--------------030204050404030606080807-- --===============8862587905135692994== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel --===============8862587905135692994==--