linux-kernel - [PATCH 2/2] MIPS: Fix R2300 FP context switch handling

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <1442855262-25416-2-git-send-email-paul.burton@imgtec.com>
Date:	Mon, 21 Sep 2015 10:07:42 -0700
From:	Paul Burton <paul.burton@...tec.com>
To:	<linux-mips@...ux-mips.org>
CC:	Paul Burton <paul.burton@...tec.com>,
	"Maciej W. Rozycki" <macro@...ux-mips.org>,
	<linux-kernel@...r.kernel.org>, Ralf Baechle <ralf@...ux-mips.org>,
	Manuel Lauss <manuel.lauss@...il.com>
Subject: [PATCH 2/2] MIPS: Fix R2300 FP context switch handling

Commit 1a3d59579b9f ("MIPS: Tidy up FPU context switching") removed FP
context saving from the asm-written resume function in favour of reusing
existing code to perform the same task. However it only removed the FP
context saving code from the r4k_switch.S implementation of resume.
Remove it from the r2300_switch.S implementation too in order to prevent
attempting to save the FP context twice, which would likely lead to an
exception from the second save because the FPU had already been disabled
by the first save.

This patch has only been build tested, using rbtx49xx_defconfig.

Fixes: 1a3d59579b9f ("MIPS: Tidy up FPU context switching")
Signed-off-by: Paul Burton <paul.burton@...tec.com>
---

 arch/mips/kernel/r2300_switch.S | 28 +---------------------------
 1 file changed, 1 insertion(+), 27 deletions(-)

diff --git a/arch/mips/kernel/r2300_switch.S b/arch/mips/kernel/r2300_switch.S
index 5087a4b7..ac27ef7 100644
--- a/arch/mips/kernel/r2300_switch.S
+++ b/arch/mips/kernel/r2300_switch.S
@@ -31,18 +31,8 @@
 #define ST_OFF (_THREAD_SIZE - 32 - PT_SIZE + PT_STATUS)
 
 /*
- * FPU context is saved iff the process has used it's FPU in the current
- * time slice as indicated by TIF_USEDFPU.  In any case, the CU1 bit for user
- * space STATUS register should be 0, so that a process *always* starts its
- * userland with FPU disabled after each context switch.
- *
- * FPU will be enabled as soon as the process accesses FPU again, through
- * do_cpu() trap.
- */
-
-/*
  * task_struct *resume(task_struct *prev, task_struct *next,
- *		       struct thread_info *next_ti, int usedfpu)
+ *		       struct thread_info *next_ti)
  */
 LEAF(resume)
 	mfc0	t1, CP0_STATUS
@@ -50,22 +40,6 @@ LEAF(resume)
 	cpu_save_nonscratch a0
 	sw	ra, THREAD_REG31(a0)
 
-	beqz	a3, 1f
-
-	PTR_L	t3, TASK_THREAD_INFO(a0)
-
-	/*
-	 * clear saved user stack CU1 bit
-	 */
-	lw	t0, ST_OFF(t3)
-	li	t1, ~ST0_CU1
-	and	t0, t0, t1
-	sw	t0, ST_OFF(t3)
-
-	fpu_save_single a0, t0			# clobbers t0
-
-1:
-
 #if defined(CONFIG_CC_STACKPROTECTOR) && !defined(CONFIG_SMP)
 	PTR_LA	t8, __stack_chk_guard
 	LONG_L	t9, TASK_STACK_CANARY(a1)
-- 
2.5.3

--
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/