summaryrefslogtreecommitdiffstats
path: root/arch/sparc/kernel/visemul.c
diff options
context:
space:
mode:
authorDavid S. Miller <davem@davemloft.net>2012-10-26 15:18:37 -0700
committerDavid S. Miller <davem@davemloft.net>2012-10-26 15:18:37 -0700
commit517ffce4e1a03aea979fe3a18a3dd1761a24fafb (patch)
treec470fe3e6266dd96c7c6f0c9df881712d89df546 /arch/sparc/kernel/visemul.c
parent1d47091ac6bf1286d708ebcd3f2b69d7c682916b (diff)
downloadlinux-517ffce4e1a03aea979fe3a18a3dd1761a24fafb.tar.gz
linux-517ffce4e1a03aea979fe3a18a3dd1761a24fafb.tar.bz2
linux-517ffce4e1a03aea979fe3a18a3dd1761a24fafb.zip
sparc64: Make montmul/montsqr/mpmul usable in 32-bit threads.
The Montgomery Multiply, Montgomery Square, and Multiple-Precision Multiply instructions work by loading a combination of the floating point and multiple register windows worth of integer registers with the inputs. These values are 64-bit. But for 32-bit userland processes we only save the low 32-bits of each integer register during a register spill. This is because the register window save area is in the user stack and has a fixed layout. Therefore, the only way to use these instruction in 32-bit mode is to perform the following sequence: 1) Load the top-32bits of a choosen integer register with a sentinel, say "-1". This will be in the outer-most register window. The idea is that we're trying to see if the outer-most register window gets spilled, and thus the 64-bit values were truncated. 2) Load all the inputs for the montmul/montsqr/mpmul instruction, down to the inner-most register window. 3) Execute the opcode. 4) Traverse back up to the outer-most register window. 5) Check the sentinel, if it's still "-1" store the results. Otherwise retry the entire sequence. This retry is extremely troublesome. If you're just unlucky and an interrupt or other trap happens, it'll push that outer-most window to the stack and clear the sentinel when we restore it. We could retry forever and never make forward progress if interrupts arrive at a fast enough rate (consider perf events as one example). So we have do limited retries and fallback to software which is extremely non-deterministic. Luckily it's very straightforward to provide a mechanism to let 32-bit applications use a 64-bit stack. Stacks in 64-bit mode are biased by 2047 bytes, which means that the lowest bit is set in the actual %sp register value. So if we see bit zero set in a 32-bit application's stack we treat it like a 64-bit stack. Runtime detection of such a facility is tricky, and cumbersome at best. For example, just trying to use a biased stack and seeing if it works is hard to recover from (the signal handler will need to use an alt stack, plus something along the lines of longjmp). Therefore, we add a system call to report a bitmask of arch specific features like this in a cheap and less hairy way. With help from Andy Polyakov. Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'arch/sparc/kernel/visemul.c')
-rw-r--r--arch/sparc/kernel/visemul.c23
1 files changed, 14 insertions, 9 deletions
diff --git a/arch/sparc/kernel/visemul.c b/arch/sparc/kernel/visemul.c
index 08e074b7eb6a..c096c624ac4d 100644
--- a/arch/sparc/kernel/visemul.c
+++ b/arch/sparc/kernel/visemul.c
@@ -149,21 +149,24 @@ static inline void maybe_flush_windows(unsigned int rs1, unsigned int rs2,
static unsigned long fetch_reg(unsigned int reg, struct pt_regs *regs)
{
- unsigned long value;
+ unsigned long value, fp;
if (reg < 16)
return (!reg ? 0 : regs->u_regs[reg]);
+
+ fp = regs->u_regs[UREG_FP];
+
if (regs->tstate & TSTATE_PRIV) {
struct reg_window *win;
- win = (struct reg_window *)(regs->u_regs[UREG_FP] + STACK_BIAS);
+ win = (struct reg_window *)(fp + STACK_BIAS);
value = win->locals[reg - 16];
- } else if (test_thread_flag(TIF_32BIT)) {
+ } else if (!test_thread_64bit_stack(fp)) {
struct reg_window32 __user *win32;
- win32 = (struct reg_window32 __user *)((unsigned long)((u32)regs->u_regs[UREG_FP]));
+ win32 = (struct reg_window32 __user *)((unsigned long)((u32)fp));
get_user(value, &win32->locals[reg - 16]);
} else {
struct reg_window __user *win;
- win = (struct reg_window __user *)(regs->u_regs[UREG_FP] + STACK_BIAS);
+ win = (struct reg_window __user *)(fp + STACK_BIAS);
get_user(value, &win->locals[reg - 16]);
}
return value;
@@ -172,16 +175,18 @@ static unsigned long fetch_reg(unsigned int reg, struct pt_regs *regs)
static inline unsigned long __user *__fetch_reg_addr_user(unsigned int reg,
struct pt_regs *regs)
{
+ unsigned long fp = regs->u_regs[UREG_FP];
+
BUG_ON(reg < 16);
BUG_ON(regs->tstate & TSTATE_PRIV);
- if (test_thread_flag(TIF_32BIT)) {
+ if (!test_thread_64bit_stack(fp)) {
struct reg_window32 __user *win32;
- win32 = (struct reg_window32 __user *)((unsigned long)((u32)regs->u_regs[UREG_FP]));
+ win32 = (struct reg_window32 __user *)((unsigned long)((u32)fp));
return (unsigned long __user *)&win32->locals[reg - 16];
} else {
struct reg_window __user *win;
- win = (struct reg_window __user *)(regs->u_regs[UREG_FP] + STACK_BIAS);
+ win = (struct reg_window __user *)(fp + STACK_BIAS);
return &win->locals[reg - 16];
}
}
@@ -204,7 +209,7 @@ static void store_reg(struct pt_regs *regs, unsigned long val, unsigned long rd)
} else {
unsigned long __user *rd_user = __fetch_reg_addr_user(rd, regs);
- if (test_thread_flag(TIF_32BIT))
+ if (!test_thread_64bit_stack(regs->u_regs[UREG_FP]))
__put_user((u32)val, (u32 __user *)rd_user);
else
__put_user(val, rd_user);