On Tue, Oct 11, 2022 at 10:28:44AM -0300, Martin Fernandez wrote:
On 10/7/22, Kirill A. Shutemov kirill@shutemov.name wrote:
On Mon, Jul 04, 2022 at 10:58:26AM -0300, Martin Fernandez wrote:
Add a new member in the pg_data_t struct to tell whether the node corresponding to that pg_data_t is able to do hardware memory encryption.
This will be read from sysfs.
Signed-off-by: Martin Fernandez martin.fernandez@eclypsium.com
include/linux/mmzone.h | 3 +++ mm/page_alloc.c | 1 + 2 files changed, 4 insertions(+)
diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index aab70355d64f..6fd4785f1d05 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -883,6 +883,9 @@ typedef struct pglist_data { struct task_struct *kcompactd; bool proactive_compact_trigger; #endif
- bool crypto_capable;
There's already pgdat->flags. Any reason we cannot encode it there?
Not really a reason, I'll considerate when I send then next version. I tried to quickly find for references of what kind of flags does it have, I didn't find any. Do you suggest it should work?
Maybe. Or maybe introduce capabilities bitfield and make crypto as one of them.
We consider to re-use the approach for other cases. Like if the memory in the node is TDX-compatible (there's more requirements for it than just encryption).