From d3ea9fa0a563620fe9f416f94bb8927c64390917 Mon Sep 17 00:00:00 2001 From: Stuart Menefy Date: Fri, 25 Sep 2009 18:25:10 +0100 Subject: sh: Minor optimisations to FPU handling A number of small optimisations to FPU handling, in particular: - move the task USEDFPU flag from the thread_info flags field (which is accessed asynchronously to the thread) to a new status field, which is only accessed by the thread itself. This allows locking to be removed in most cases, or can be reduced to a preempt_lock(). This mimics the i386 behaviour. - move the modification of regs->sr and thread_info->status flags out of save_fpu() to __unlazy_fpu(). This gives the compiler a better chance to optimise things, as well as making save_fpu() symmetrical with restore_fpu() and init_fpu(). - implement prepare_to_copy(), so that when creating a thread, we can unlazy the FPU prior to copying the thread data structures. Also make sure that the FPU is disabled while in the kernel, in particular while booting, and for newly created kernel threads, In a very artificial benchmark, the execution time for 2500000 context switches was reduced from 50 to 45 seconds. Signed-off-by: Stuart Menefy Signed-off-by: Paul Mundt --- arch/sh/math-emu/math.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'arch/sh/math-emu') diff --git a/arch/sh/math-emu/math.c b/arch/sh/math-emu/math.c index ac2d7abd2567..d6c15cae0912 100644 --- a/arch/sh/math-emu/math.c +++ b/arch/sh/math-emu/math.c @@ -558,7 +558,7 @@ static int ieee_fpe_handler(struct pt_regs *regs) (finsn >> 8) & 0xf); tsk->thread.fpu.hard.fpscr &= ~(FPSCR_CAUSE_MASK | FPSCR_FLAG_MASK); - set_tsk_thread_flag(tsk, TIF_USEDFPU); + task_thread_info(tsk)->status |= TS_USEDFPU; } else { info.si_signo = SIGFPE; info.si_errno = 0; @@ -619,10 +619,10 @@ int do_fpu_inst(unsigned short inst, struct pt_regs *regs) struct task_struct *tsk = current; struct sh_fpu_soft_struct *fpu = &(tsk->thread.fpu.soft); - if (!test_tsk_thread_flag(tsk, TIF_USEDFPU)) { + if (!(task_thread_info(tsk)->status & TS_USEDFPU)) { /* initialize once. */ fpu_init(fpu); - set_tsk_thread_flag(tsk, TIF_USEDFPU); + task_thread_info(tsk)->status |= TS_USEDFPU; } return fpu_emulate(inst, fpu, regs); -- cgit v1.2.2 From 0ea820cf9bf58f735ed40ec67947159c4f170012 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Wed, 13 Jan 2010 12:51:40 +0900 Subject: sh: Move over to dynamically allocated FPU context. This follows the x86 xstate changes and implements a task_xstate slab cache that is dynamically sized to match one of hard FP/soft FP/FPU-less. This also tidies up and consolidates some of the SH-2A/SH-4 FPU fragmentation. Now fpu state restorers are commonly defined, with the init_fpu()/fpu_init() mess reworked to follow the x86 convention. The fpu_init() register initialization has been replaced by xstate setup followed by writing out to hardware via the standard restore path. As init_fpu() now performs a slab allocation a secondary lighterweight restorer is also introduced for the context switch. In the future the DSP state will be rolled in here, too. More work remains for math emulation and the SH-5 FPU, which presently uses its own special (UP-only) interfaces. Signed-off-by: Paul Mundt --- arch/sh/math-emu/math.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'arch/sh/math-emu') diff --git a/arch/sh/math-emu/math.c b/arch/sh/math-emu/math.c index d6c15cae0912..1fcdb1220975 100644 --- a/arch/sh/math-emu/math.c +++ b/arch/sh/math-emu/math.c @@ -471,10 +471,10 @@ static int fpu_emulate(u16 code, struct sh_fpu_soft_struct *fregs, struct pt_reg * denormal_to_double - Given denormalized float number, * store double float * - * @fpu: Pointer to sh_fpu_hard structure + * @fpu: Pointer to sh_fpu_soft structure * @n: Index to FP register */ -static void denormal_to_double(struct sh_fpu_hard_struct *fpu, int n) +static void denormal_to_double(struct sh_fpu_soft_struct *fpu, int n) { unsigned long du, dl; unsigned long x = fpu->fpul; @@ -552,11 +552,11 @@ static int ieee_fpe_handler(struct pt_regs *regs) if ((finsn & 0xf1ff) == 0xf0ad) { /* fcnvsd */ struct task_struct *tsk = current; - if ((tsk->thread.fpu.hard.fpscr & (1 << 17))) { + if ((tsk->thread.xstate->softfpu.fpscr & (1 << 17))) { /* FPU error */ - denormal_to_double (&tsk->thread.fpu.hard, + denormal_to_double (&tsk->thread.xstate->softfpu, (finsn >> 8) & 0xf); - tsk->thread.fpu.hard.fpscr &= + tsk->thread.xstate->softfpu.fpscr &= ~(FPSCR_CAUSE_MASK | FPSCR_FLAG_MASK); task_thread_info(tsk)->status |= TS_USEDFPU; } else { @@ -617,7 +617,7 @@ static void fpu_init(struct sh_fpu_soft_struct *fpu) int do_fpu_inst(unsigned short inst, struct pt_regs *regs) { struct task_struct *tsk = current; - struct sh_fpu_soft_struct *fpu = &(tsk->thread.fpu.soft); + struct sh_fpu_soft_struct *fpu = &(tsk->thread.xstate->softfpu); if (!(task_thread_info(tsk)->status & TS_USEDFPU)) { /* initialize once. */ -- cgit v1.2.2