Escalated to Bugzilla from IssueTracker
Reposting the patch publicly.
Excerpt from ChangeLog-2.6.14-rc1:
When copying data from user-space to kernel-space by __copy_user(),
a page_not_present fault sometimes occurs at vmalloced kernel address
because of VHPT pre-fetching.
Ignore the page_not_present fault in ia64_do_page_fault() before
jumping into exception handlers.
diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c
@@ -206,9 +206,6 @@ ia64_do_page_fault (unsigned long addres
- if (done_with_exception(regs))
* Since we have no vma's for region 5, we might get here even if the
* valid, due to the VHPT walker inserting a non present translation that
@@ -219,6 +216,9 @@ ia64_do_page_fault (unsigned long addres
if (REGION_NUMBER(address) == 5 && mapped_kernel_page_is_present(address))
+ if (done_with_exception(regs))
* Oops. The kernel tried to access some bad page. We'll have to
* with extreme prejudice.
In answer to comment #10, the fix is queued for the next U7 build, which
will probably occur tomorrow night.
A fix for this problem has just been committed to the RHEL3 U7
patch pool this evening (in kernel version 2.4.21-37.12.EL).
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.