]> Git Repo - linux.git/commitdiff
x86/speculation/l1tf: Suggest what to do on systems with too much RAM
authorVlastimil Babka <[email protected]>
Thu, 23 Aug 2018 14:21:29 +0000 (16:21 +0200)
committerThomas Gleixner <[email protected]>
Fri, 24 Aug 2018 13:55:17 +0000 (15:55 +0200)
Two users have reported [1] that they have an "extremely unlikely" system
with more than MAX_PA/2 memory and L1TF mitigation is not effective.

Make the warning more helpful by suggesting the proper mem=X kernel boot
parameter to make it effective and a link to the L1TF document to help
decide if the mitigation is worth the unusable RAM.

[1] https://bugzilla.suse.com/show_bug.cgi?id=1105536

Suggested-by: Michal Hocko <[email protected]>
Signed-off-by: Vlastimil Babka <[email protected]>
Acked-by: Michal Hocko <[email protected]>
Cc: "H . Peter Anvin" <[email protected]>
Cc: Linus Torvalds <[email protected]>
Cc: Andi Kleen <[email protected]>
Cc: Dave Hansen <[email protected]>
Cc: [email protected]
Link: https://lkml.kernel.org/r/[email protected]
arch/x86/kernel/cpu/bugs.c

index cb4a16292aa7cb84c21d175844f3122db016a0a0..4c2313d0b9caadb71cd995ef584740b3113af0c1 100644 (file)
@@ -702,6 +702,10 @@ static void __init l1tf_select_mitigation(void)
        half_pa = (u64)l1tf_pfn_limit() << PAGE_SHIFT;
        if (e820__mapped_any(half_pa, ULLONG_MAX - half_pa, E820_TYPE_RAM)) {
                pr_warn("System has more than MAX_PA/2 memory. L1TF mitigation not effective.\n");
+               pr_info("You may make it effective by booting the kernel with mem=%llu parameter.\n",
+                               half_pa);
+               pr_info("However, doing so will make a part of your RAM unusable.\n");
+               pr_info("Reading https://www.kernel.org/doc/html/latest/admin-guide/l1tf.html might help you decide.\n");
                return;
        }
 
This page took 0.051714 seconds and 4 git commands to generate.