tile: fast-path unaligned memory access for tilegx

This change enables unaligned userspace memory access via a kernel
fast path on tilegx.  The kernel tracks user PC/instruction pairs
per-thread using a direct-mapped cache in userspace.  The cache
maps those PC/instruction pairs to JIT'ed instruction sequences that
load or store using byte-wide load store intructions and then
synthesize 2-, 4- or 8-byte load or store results.  Once an
instruction has been seen to generate an unaligned access once,
subsequent hits on that instruction typically require overhead
of only around 50 cycles if cache and TLB is hot.

We support the prctl() PR_GET_UNALIGN / PR_SET_UNALIGN sys call to
enable or disable unaligned fixups on a per-process basis.

To do this we pull some of the tilepro unaligned support out of the
single_step.c file; tilepro uses instruction disassembly for both
single-step and unaligned access support.  Since tilegx actually has
hardware singlestep support, though, it's cleaner to keep the tilegx
unaligned access code in a separate file.  While we're at it,
properly rename the tilepro-specific types, etc., to have tilepro
suffixes instead of generic tile suffixes.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>

1 files changed, 21 insertions, 0 deletions

diff --git a/arch/tile/kernel/process.c b/arch/tile/kernel/process.c
index 8d6c51d55762..25678b83b747 100644
--- a/arch/tile/kernel/process.c
+++ b/arch/tile/kernel/process.c
@@ -33,6 +33,7 @@
 #include <asm/syscalls.h>
 #include <asm/traps.h>
 #include <asm/setup.h>
+#include <asm/uaccess.h>
 #ifdef CONFIG_HARDWALL
 #include <asm/hardwall.h>
 #endif
@@ -147,6 +148,14 @@ int copy_thread(unsigned long clone_flags, unsigned long sp,
 	 */
 	task_thread_info(p)->step_state = NULL;
 
+#ifdef __tilegx__
+	/*
+	 * Do not clone unalign jit fixup from the parent; each thread
+	 * must allocate its own on demand.
+	 */
+	task_thread_info(p)->unalign_jit_base = NULL;
+#endif
+
 	/*
 	 * Copy the registers onto the kernel stack so the
 	 * return-from-interrupt code will reload it into registers.
@@ -205,6 +214,18 @@ int copy_thread(unsigned long clone_flags, unsigned long sp,
 	return 0;
 }
 
+int set_unalign_ctl(struct task_struct *tsk, unsigned int val)
+{
+	task_thread_info(tsk)->align_ctl = val;
+	return 0;
+}
+
+int get_unalign_ctl(struct task_struct *tsk, unsigned long adr)
+{
+	return put_user(task_thread_info(tsk)->align_ctl,
+			(unsigned int __user *)adr);
+}
+
 /*
  * Return "current" if it looks plausible, or else a pointer to a dummy.
  * This can be helpful if we are just trying to emit a clean panic.