| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Sat, 10 Oct 2015 18:41:40 +0200
From: "Jason A. Donenfeld" <Jason@...c4.com>
To: stable@...r.kernel.org
Cc: linux-kernel@...r.kernel.org, yanirx.lubetkin@...el.com,
raanan.avargil@...el.com, jeffrey.t.kirsher@...el.com
Subject: Re: e1000e: hard system lockup on Linux 4.2
On Fri, Oct 9, 2015 at 9:21 PM, Jason A. Donenfeld <Jason@...c4.com> wrote:
> In the meantime while I wait to hear back, I'll try backporting that
> commit to 4.2 myself, and seeing the stability of my laptop over the
> next 24 hours.
I can confirm that the fix was successful.
Stable - please add 37b12910dd11d9ab969f2c310dc9160b7f3e3405 to the
4.2.4 release.
=~=~=~=~=~=~=~=
>From 37b12910dd11d9ab969f2c310dc9160b7f3e3405 Mon Sep 17 00:00:00 2001
From: Raanan Avargil <raanan.avargil@...el.com>
Date: Sun, 19 Jul 2015 16:33:20 +0300
Subject: [PATCH] e1000e: Fix tight loop implementation of systime read
algorithm
Change the algorithm. Read systimel twice and check for overflow.
If there was no overflow, use the first value.
If there was an overflow, read systimeh again and use the second
systimel value.
Signed-off-by: Raanan Avargil <raanan.avargil@...el.com>
Tested-by: Aaron Brown <aaron.f.brown@...el.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@...el.com>
---
drivers/net/ethernet/intel/e1000e/netdev.c | 31 ++++++++++++++++++++----------
1 file changed, 21 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c
b/drivers/net/ethernet/intel/e1000e/netdev.c
index 24b7269..96a8166 100644
--- a/drivers/net/ethernet/intel/e1000e/netdev.c
+++ b/drivers/net/ethernet/intel/e1000e/netdev.c
@@ -4280,18 +4280,29 @@ static cycle_t e1000e_cyclecounter_read(const
struct cyclecounter *cc)
struct e1000_adapter *adapter = container_of(cc, struct e1000_adapter,
cc);
struct e1000_hw *hw = &adapter->hw;
+ u32 systimel_1, systimel_2, systimeh;
cycle_t systim, systim_next;
- /* SYSTIMH latching upon SYSTIML read does not work well. To fix that
- * we don't want to allow overflow of SYSTIML and a change to SYSTIMH
- * to occur between reads, so if we read a vale close to overflow, we
- * wait for overflow to occur and read both registers when its safe.
+ /* SYSTIMH latching upon SYSTIML read does not work well.
+ * This means that if SYSTIML overflows after we read it but before
+ * we read SYSTIMH, the value of SYSTIMH has been incremented and we
+ * will experience a huge non linear increment in the systime value
+ * to fix that we test for overflow and if true, we re-read systime.
*/
- u32 systim_overflow_latch_fix = 0x3FFFFFFF;
-
- do {
- systim = (cycle_t)er32(SYSTIML);
- } while (systim > systim_overflow_latch_fix);
- systim |= (cycle_t)er32(SYSTIMH) << 32;
+ systimel_1 = er32(SYSTIML);
+ systimeh = er32(SYSTIMH);
+ systimel_2 = er32(SYSTIML);
+ /* Check for overflow. If there was no overflow, use the values */
+ if (systimel_1 < systimel_2) {
+ systim = (cycle_t)systimel_1;
+ systim |= (cycle_t)systimeh << 32;
+ } else {
+ /* There was an overflow, read again SYSTIMH, and use
+ * systimel_2
+ */
+ systimeh = er32(SYSTIMH);
+ systim = (cycle_t)systimel_2;
+ systim |= (cycle_t)systimeh << 32;
+ }
if ((hw->mac.type == e1000_82574) || (hw->mac.type == e1000_82583)) {
u64 incvalue, time_delta, rem, temp;
--
2.6.0
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists