[Bug testsuite/15928] New: the memory1.exp testcase is accessing user memory as kernel memory

[Bug testsuite/15928] New: the memory1.exp testcase is accessing user memory as kernel memory

[dsmith at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=15928

            Bug ID: 15928
           Summary: the memory1.exp testcase is accessing user memory as
                    kernel memory
           Product: systemtap
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: testsuite
          Assignee: systemtap at sourceware dot org
          Reporter: dsmith at redhat dot com

The testsuite/systemtap.printf/memory1.exp testcase has an associated
testsuite/systemtap.printf/memory1.stp file, which does this:

====
probe syscall.open {
        ...
        result = sprintf ("%m", $filename);
        ...
====

However, since $filename is a pointer to user memory, this really isn't valid.
This occidentally works on systems where there is no real user/kernel memory
split, like on x86_64, but fails on systems like s390x where there is a real
user/kernel memory split.

As the documentation for '%m' states:

              %m     Safely reads kernel memory at the given address, outputs
                     its content.  The precision specifier determines the num‐
                     ber of bytes to read.  Default is 1 byte.

The testcase will need to be fixed, perhaps reading data from a private kernel
module instead.

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

[Bug testsuite/15928] the memory1.exp testcase is accessing user memory as kernel memory

[dsmith at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=15928

David Smith <dsmith at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #1 from David Smith <dsmith at redhat dot com> ---
Fixed in commit 0447e0e by probing a function in a new kernel module that takes
kernel memory.

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