Skip to contentSkip to conversation

m68k: make support for FPU hardware configurable

PATCH In review [email protected] 2 participants 5 messages Feb 10, 2012 – Feb 21, 2012

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/2 m68k: make support for FPU hardware configurable

From: Greg Ungerer <[email protected]>

The classic m68k code has always supported an FPU (although it may have been a software emulated one). The non-MMU m68k code has never supported FPU hardware. To help in merging common code create a configation setting that signifies if we are builing in FPU support or not.

This switch, CONFIG_FPU, is set as per the current use cases. So it is always enabled if CONFIG_MMU is set, and disabled otherwise. With a little extra code it will be possible to disable it on the classic m68k platforms as well, and to enable it on non-MMU platforms that do have hardware FPU.

---
 arch/m68k/Kconfig |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)
arch/m68k/Kconfig +4 -0
index ae413d4..57bf848 100644
@@ -7,6 +7,7 @@ config M68K
 	select GENERIC_IRQ_SHOW
 	select ARCH_HAVE_NMI_SAFE_CMPXCHG if RMW_INSNS
 	select GENERIC_CPU_DEVICES
+	select FPU if MMU
 
 config RWSEM_GENERIC_SPINLOCK
 	bool
@@ -67,6 +68,9 @@ config CPU_HAS_NO_MULDIV64
 config CPU_HAS_ADDRESS_SPACES
 	bool
 
+config FPU
+	bool
+
 config HZ
 	int
 	default 1000 if CLEOPATRA
Signature · 1 line
1.7.0.4
#1 lore raw

PATCH 2/2 m68k: merge the MMU and non-MMU versions of process.c

From: Greg Ungerer <[email protected]>

The MMU and non-MMU varients of the m68k arch process.c code are pretty much the same. Only a few minor details differ between the two. The majority of the difference is to deal with having or wanting hardware FPU support. So merge them back into a single process.c file.

---
 arch/m68k/kernel/{process_mm.c => process.c} |   19 +-
 arch/m68k/kernel/process_no.c                |  406 --------------------------
 2 files changed, 15 insertions(+), 410 deletions(-)
 rename arch/m68k/kernel/{process_mm.c => process.c} (97%)
 delete mode 100644 arch/m68k/kernel/process_no.c
arch/m68k/kernel/process.c renamed from arch/m68k/kernel/process_mm.c +15 -4
similarity index 97%
rename from arch/m68k/kernel/process_mm.c
rename to arch/m68k/kernel/process.c
index 099283e..f053e24 100644
@@ -144,8 +144,10 @@ int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
 	   "trap #0\n\t"		/* Linux/m68k system call */
 	   "tstl %0\n\t"		/* child or parent */
 	   "jne 1f\n\t"			/* parent - jump */
+#ifdef CONFIG_MMU
 	   "lea %%sp@(%c7),%6\n\t"	/* reload current */
 	   "movel %6@,%6\n\t"
+#endif
 	   "movel %3,%%sp@-\n\t"	/* push argument */
 	   "jsr %4@\n\t"		/* call fn */
 	   "movel %0,%%d1\n\t"		/* pass exit value */
@@ -168,11 +170,13 @@ EXPORT_SYMBOL(kernel_thread);
 
 void flush_thread(void)
 {
-	unsigned long zero = 0;
-
 	current->thread.fs = __USER_DS;
-	if (!FPU_IS_EMU)
+#ifdef CONFIG_FPU
+	if (!FPU_IS_EMU) {
+		unsigned long zero = 0;
 		asm volatile("frestore %0": :"m" (zero));
+	}
+#endif
 }
 
 /*
@@ -184,7 +188,11 @@ void flush_thread(void)
 
 asmlinkage int m68k_fork(struct pt_regs *regs)
 {
+#ifdef CONFIG_MMU
 	return do_fork(SIGCHLD, rdusp(), regs, 0, NULL, NULL);
+#else
+	return -EINVAL;
+#endif
 }
 
 asmlinkage int m68k_vfork(struct pt_regs *regs)
@@ -242,6 +250,7 @@ int copy_thread(unsigned long clone_flags, unsigned long usp,
 	 */
 	p->thread.fs = get_fs().seg;
 
+#ifdef CONFIG_FPU
 	if (!FPU_IS_EMU) {
 		/* Copy the current fpu state */
 		asm volatile ("fsave %0" : : "m" (p->thread.fpstate[0]) : "memory");
@@ -271,12 +280,13 @@ int copy_thread(unsigned long clone_flags, unsigned long usp,
 		/* Restore the state in case the fpu was busy */
 		asm volatile ("frestore %0" : : "m" (p->thread.fpstate[0]));
 	}
+#endif /* CONFIG_FPU */
 
 	return 0;
 }
 
 /* Fill in the fpu structure for a core dump.  */
-
+#ifdef CONFIG_FPU
 int dump_fpu (struct pt_regs *regs, struct user_m68kfp_struct *fpu)
 {
 	char fpustate[216];
@@ -325,6 +335,7 @@ int dump_fpu (struct pt_regs *regs, struct user_m68kfp_struct *fpu)
 	return 1;
 }
 EXPORT_SYMBOL(dump_fpu);
+#endif /* CONFIG_FPU */
 
 /*
  * sys_execve() executes a new program.
arch/m68k/kernel/process_no.c deleted +0 -406

406 lines not on this page · raw message

Signature · 1 line
1.7.0.4
maintainer Acked-by on v1 2/2
#2 lore raw

Re: [PATCH 2/2] m68k: merge the MMU and non-MMU versions of process.c

Quoted 6 lines · On Fri, Feb 10, 2012 at 07:14, <[email protected]> wrote:

The MMU and non-MMU varients of the m68k arch process.c code are pretty much the same. Only a few minor details differ between the two. The majority of the difference is to deal with having or wanting hardware FPU support. So merge them back into a single process.c file.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds

maintainer Acked-by on v1 1/2
#3 lore raw
In reply to [email protected]
Quoted 11 lines · On Fri, Feb 10, 2012 at 07:14, <[email protected]> wrote:

The classic m68k code has always supported an FPU (although it may have been a software emulated one). The non-MMU m68k code has never supported FPU hardware. To help in merging common code create a configation setting that signifies if we are builing in FPU support or not.

This switch, CONFIG_FPU, is set as per the current use cases. So it is always enabled if CONFIG_MMU is set, and disabled otherwise. With a little extra code it will be possible to disable it on the classic m68k platforms as well, and to enable it on non-MMU platforms that do have hardware FPU.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds

#4 lore raw

Hi Geert,

Quoted 6 lines · Geert Uytterhoeven
Quoted 5 lines · On Fri, Feb 10, 2012 at 07:14,<[email protected]> wrote:

The classic m68k code has always supported an FPU (although it may have been a software emulated one). The non-MMU m68k code has never supported FPU hardware. To help in merging common code create a configation setting that signifies if we are builing in FPU support or not.

Quoted 8 lines
Quoted 6 lines

This switch, CONFIG_FPU, is set as per the current use cases. So it is always enabled if CONFIG_MMU is set, and disabled otherwise. With a little extra code it will be possible to disable it on the classic m68k platforms as well, and to enable it on non-MMU platforms that do have hardware FPU.

Thanks!

Regards
Greg

------------------------------------------------------------------------

Greg Ungerer  --  Principal Engineer        EMAIL:     [email protected]
SnapGear Group, McAfee                      PHONE:       +61 7 3435 2888
8 Gardner Close                             FAX:         +61 7 3217 5323
Milton, QLD, 4064, Australia                WEB: http://www.SnapGear.com