x86/speculation/l1tf: Drop the swap storage limit restriction when l1tf=off
authorMichal Hocko <mhocko@suse.com>
Tue, 13 Nov 2018 18:49:10 +0000 (19:49 +0100)
committerThomas Gleixner <tglx@linutronix.de>
Tue, 11 Dec 2018 10:46:13 +0000 (11:46 +0100)
Swap storage is restricted to max_swapfile_size (~16TB on x86_64) whenever
the system is deemed affected by L1TF vulnerability. Even though the limit
is quite high for most deployments it seems to be too restrictive for
deployments which are willing to live with the mitigation disabled.

We have a customer to deploy 8x 6,4TB PCIe/NVMe SSD swap devices which is
clearly out of the limit.

Drop the swap restriction when l1tf=off is specified. It also doesn't make
much sense to warn about too much memory for the l1tf mitigation when it is
forcefully disabled by the administrator.

[ tglx: Folded the documentation delta change ]

Fixes: 377eeaa8e11f ("x86/speculation/l1tf: Limit swap file size to MAX_PA/2")
Signed-off-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Pavel Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Kosina <jkosina@suse.cz>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: <linux-mm@kvack.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181113184910.26697-1-mhocko@kernel.org

index 05a252e5178d7af82a0541cc50bd178218afde10..835e422572eb9c6b507c9e56e5f4c29c731582b0 100644 (file)
                                Disables hypervisor mitigations and doesn't
                                emit any warnings.
+                               It also drops the swap size and available
+                               RAM limit restriction on both hypervisor and
+                               bare metal.
                        Default is 'flush'.
index b85dd80510b0262382c7a41890da5534ad09da03..9af977384168eacd8cbe621615482e782b7b05c9 100644 (file)
@@ -405,6 +405,9 @@ time with the option "l1tf=". The valid arguments for this option are:
   off          Disables hypervisor mitigations and doesn't emit any
+               It also drops the swap size and available RAM limit restrictions
+               on both hypervisor and bare metal.
   ============  =============================================================
 The default is 'flush'. For details about L1D flushing see :ref:`l1d_flush`.
@@ -576,7 +579,8 @@ Default mitigations
   The kernel default mitigations for vulnerable processors are:
   - PTE inversion to protect against malicious user space. This is done
-    unconditionally and cannot be controlled.
+    unconditionally and cannot be controlled. The swap storage is limited
+    to ~16TB.
   - L1D conditional flushing on VMENTER when EPT is enabled for
     a guest.
index a68b32cb845a4478286144578e811de85844fba9..58689ac64440d0657f87d76b93244b63e88cb723 100644 (file)
@@ -1002,7 +1002,8 @@ 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)) {
+       if (l1tf_mitigation != L1TF_MITIGATION_OFF &&
+                       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",
index ef99f3892e1faedbf96a935838c248c4801dbc2b..427a955a2cf2c49fdc1fd5c84494f064a5704caf 100644 (file)
@@ -931,7 +931,7 @@ unsigned long max_swapfile_size(void)
        pages = generic_max_swapfile_size();
-       if (boot_cpu_has_bug(X86_BUG_L1TF)) {
+       if (boot_cpu_has_bug(X86_BUG_L1TF) && l1tf_mitigation != L1TF_MITIGATION_OFF) {
                /* Limit the swap file size to MAX_PA/2 for L1TF workaround */
                unsigned long long l1tf_limit = l1tf_pfn_limit();