Jeff Moyer <[email protected]> writes:
> Matthew Wilcox <[email protected]> writes:
>
>> This custom resizing array was vulnerable to a Spectre attack (speculating
>> off the end of an array to a user-controlled offset). The XArray is
>> not vulnerable to Spectre as it always masks its lookups to be within
>> the bounds of the array.
>
> I'm not a big fan of completely re-writing the code to fix this. Isn't
> the below patch sufficient?
Too quick on the draw. Here's a patch that compiles. ;-)
Cheers,
Jeff
diff --git a/fs/aio.c b/fs/aio.c
index 97f983592925..aac9659381d2 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -45,6 +45,7 @@
#include <asm/kmap_types.h>
#include <linux/uaccess.h>
+#include <linux/nospec.h>
#include "internal.h"
@@ -1038,6 +1039,7 @@ static struct kioctx *lookup_ioctx(unsigned long ctx_id)
if (!table || id >= table->nr)
goto out;
+ id = array_index_nospec(id, table->nr);
ctx = rcu_dereference(table->table[id]);
if (ctx && ctx->user_id == ctx_id) {
if (percpu_ref_tryget_live(&ctx->users))
