[OpenAFS] Re: More Linux 2.6.10 -- page_follow_link test typo

[Matthew Miller]

> Shouldn't that last "page_follow_page" be "page_follow_link" instead?

Oh, wait, it's worse than that, because just having a function which doesn't
exist doesn't cause an error by default, meaning the test always succeeds. A
simple fix (plus the typo fix) follows. And with this, and with the
TASK_ZOMBIE block -- actually, the entire body of the
afs_osi_TraverseProcTable() function -- zapped, everything seems to build
and run.

But it occurs to me that this test problem probably affects many of the
other tests as well, and they probably all need this fix.

Or am I missing something here? I very well could be.


--- openafs-1.3.77/src/cf/linux-test3.m4.pagefollowlinkfix	2005-01-04 15:53:59.876475395 -0500
+++ openafs-1.3.77/src/cf/linux-test3.m4	2005-01-04 15:56:07.104503809 -0500
@@ -116,7 +116,7 @@
 AC_DEFUN([LINUX_KERNEL_PAGE_FOLLOW_LINK],[
 AC_MSG_CHECKING(for page_follow_link_light vs page_follow_link)
 save_CPPFLAGS="$CPPFLAGS"
-CPPFLAGS="-I${LINUX_KERNEL_PATH}/include -D__KERNEL__ $CPPFLAGS"
+CPPFLAGS="-Werror-implicit-function-declaration -I${LINUX_KERNEL_PATH}/include -D__KERNEL__ $CPPFLAGS"
 AC_CACHE_VAL(ac_cv_linux_kernel_page_follow_link,
 [
 AC_TRY_COMPILE(
@@ -126,5 +126,5 @@
   ],
   ac_cv_linux_kernel_page_follow_link=yes,
   ac_cv_linux_kernel_page_follow_link=no)])
-AC_MSG_RESULT($ac_cv_linux_kernel_page_follow_page)
+AC_MSG_RESULT($ac_cv_linux_kernel_page_follow_link)
 CPPFLAGS="$save_CPPFLAGS"])


-- 
Matthew Miller           mattdm@mattdm.org        <http://www.mattdm.org/>
Boston University Linux      ------>                <http://linux.bu.edu/>