On Wed, 21 Mar 2012 18:07:41 -0400
Paul Gortmaker [off-list ref] wrote:

On Mon, Mar 12, 2012 at 6:30 PM, Naoya Horiguchi
[off-list ref] wrote:

quoted

These macros will be used in later patch, where all usage are expected
to be optimized away without #ifdef CONFIG_TRANSPARENT_HUGEPAGE.
But to detect unexpected usages, we convert existing BUG() to BUILD_BUG().

Just a heads up that this showed up in linux-next today as the
cause of a new build failure for an ARM board:

http://kisskb.ellerman.id.au/kisskb/buildresult/5930053/

The internet started working again.

mm/pgtable-generic.c: In function 'pmdp_clear_flush_young':
mm/pgtable-generic.c:76: error: call to '__build_bug_failed' declared with attribute error: BUILD_BUG failed

I guess we shouldn't be evaluating HPAGE_PMD_MASK at all if
!CONFIG_TRANSPARENT_HUGEPAGE, so...

--- a/mm/pgtable-generic.c~thp-add-hpage_pmd_-definitions-for-config_transparent_hugepage-fix
+++ a/mm/pgtable-generic.c

@@ -70,10 +70,11 @@ int pmdp_clear_flush_young(struct vm_are
 			   unsigned long address, pmd_t *pmdp)
 {
 	int young;
-#ifndef CONFIG_TRANSPARENT_HUGEPAGE
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+	VM_BUG_ON(address & ~HPAGE_PMD_MASK);
+#else
 	BUG();
 #endif /* CONFIG_TRANSPARENT_HUGEPAGE */
-	VM_BUG_ON(address & ~HPAGE_PMD_MASK);
 	young = pmdp_test_and_clear_young(vma, address, pmdp);
 	if (young)
 		flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE);

_