Commit 37cd8ed9 authored by Gerhard Stenzel's avatar Gerhard Stenzel Committed by Benjamin Herrenschmidt

powerpc/cell: Make ptcal more reliable

There have been a series of checkstops on QS21 related to
ptcal being set up incorrectly. On systems that only
have memory on a single node, ptcal fails when it gets
a pointer to memory on the remote node.

Moreover, agressive prefetching in memcpy and other
functions may accidentally touch the first cache line
of the page that we reserve for ptcal, which causes
an ECC checkstop.

We now allocate pages only from the specified node, moves the
ptcal area into the middle of the allocated page to avoid
potential prefetch problems and prints the address of the
ptcal area to facilitate diagnostics.
Signed-off-by: default avatarGerhard Stenzel <gerhard.stenzel@de.ibm.com>
Signed-off-by: default avatarArnd Bergmann <arnd@arndb.de>
Signed-off-by: default avatarBenjamin Herrenschmidt <benh@kernel.crashing.org>
parent 49a84965
...@@ -122,12 +122,23 @@ static int __init cbe_ptcal_enable_on_node(int nid, int order) ...@@ -122,12 +122,23 @@ static int __init cbe_ptcal_enable_on_node(int nid, int order)
area->nid = nid; area->nid = nid;
area->order = order; area->order = order;
area->pages = alloc_pages_node(area->nid, GFP_KERNEL, area->order); area->pages = alloc_pages_node(area->nid, GFP_KERNEL | GFP_THISNODE,
area->order);
if (!area->pages) if (!area->pages) {
printk(KERN_WARNING "%s: no page on node %d\n",
__func__, area->nid);
goto out_free_area; goto out_free_area;
}
addr = __pa(page_address(area->pages)); /*
* We move the ptcal area to the middle of the allocated
* page, in order to avoid prefetches in memcpy and similar
* functions stepping on it.
*/
addr = __pa(page_address(area->pages)) + (PAGE_SIZE >> 1);
printk(KERN_DEBUG "%s: enabling PTCAL on node %d address=0x%016lx\n",
__func__, area->nid, addr);
ret = -EIO; ret = -EIO;
if (rtas_call(ptcal_start_tok, 3, 1, NULL, area->nid, if (rtas_call(ptcal_start_tok, 3, 1, NULL, area->nid,
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment