[PATCH] Another x86-64 merge - make it boot again

This fixes a few x86-64 issues that have crept in and broke bootup.

 - Disable -funit-at-a-time. It breaks symbol exports with gcc 3.3.1-hammer.
 - Fix sched_clock to not access HPET.  The new scheduler uses it more
   extensively and it is not mapped early enough.  I opted to just
   disable the HPET access, because even a slightly non monotonous TSC
   should be accurate enough for scheduling purposes. 
 - Fixes for separate objdirs by Arnd Bergmann
 - Fix two warnings that have crept in

arch/x86_64/Makefile

@@ -52,14 +52,14 @@ CFLAGS += -Wno-sign-compare
 ifneq ($(CONFIG_X86_REMOTE_DEBUG),y)
 CFLAGS += -fno-asynchronous-unwind-tables
 endif
-CFLAGS += $(call check_gcc,-funit-at-a-time,)
+#CFLAGS += $(call check_gcc,-funit-at-a-time,)
 
 head-y := arch/x86_64/kernel/head.o arch/x86_64/kernel/head64.o arch/x86_64/kernel/init_task.o
 
 libs-y 					+= arch/x86_64/lib/
 core-y					+= arch/x86_64/kernel/ arch/x86_64/mm/
 core-$(CONFIG_IA32_EMULATION)		+= arch/x86_64/ia32/
-drivers-$(CONFIG_PCI)			+= arch/x86_64/pci/
+drivers-$(CONFIG_PCI)			+= arch/i386/pci/
 drivers-$(CONFIG_OPROFILE)		+= arch/x86_64/oprofile/
 
 boot := arch/x86_64/boot

arch/x86_64/boot/Makefile

@@ -98,4 +98,4 @@ zlilo: $(BOOTIMAGE)
 	if [ -x /sbin/lilo ]; then /sbin/lilo; else /etc/lilo/install; fi
 
 install: $(BOOTIMAGE)
-	sh $(src)/install.sh $(KERNELRELEASE) $(BOOTIMAGE) System.map "$(INSTALL_PATH)"
+	sh $(srctree)/$(src)/install.sh $(KERNELRELEASE) $(BOOTIMAGE) System.map "$(INSTALL_PATH)"

arch/x86_64/kernel/Makefile

@@ -9,24 +9,19 @@ obj-y	:= process.o semaphore.o signal.o entry.o traps.o irq.o \
 		x8664_ksyms.o i387.o syscall.o vsyscall.o \
 		setup64.o bluesmoke.o bootflag.o e820.o reboot.o warmreboot.o
 
-obj-$(CONFIG_MTRR)	+= mtrr/
-obj-$(CONFIG_ACPI)	+= acpi/
-obj-$(CONFIG_X86_MSR)	+= msr.o
-obj-$(CONFIG_X86_CPUID)	+= cpuid.o
-obj-$(CONFIG_SMP)	+= smp.o smpboot.o trampoline.o
+obj-$(CONFIG_MTRR)		+= ../../i386/kernel/cpu/mtrr/
+obj-$(CONFIG_ACPI)		+= acpi/
+obj-$(CONFIG_X86_MSR)		+= msr.o
+obj-$(CONFIG_X86_CPUID)		+= cpuid.o
+obj-$(CONFIG_SMP)		+= smp.o smpboot.o trampoline.o
 obj-$(CONFIG_X86_LOCAL_APIC)	+= apic.o  nmi.o
 obj-$(CONFIG_X86_IO_APIC)	+= io_apic.o mpparse.o
 obj-$(CONFIG_SOFTWARE_SUSPEND)	+= suspend.o suspend_asm.o
-obj-$(CONFIG_EARLY_PRINTK)    += early_printk.o
-obj-$(CONFIG_GART_IOMMU) += pci-gart.o aperture.o
-obj-$(CONFIG_DUMMY_IOMMU) += pci-nommu.o pci-dma.o
+obj-$(CONFIG_EARLY_PRINTK)	+= early_printk.o
+obj-$(CONFIG_GART_IOMMU)	+= pci-gart.o aperture.o
+obj-$(CONFIG_DUMMY_IOMMU)	+= pci-nommu.o pci-dma.o
 
-obj-$(CONFIG_MODULES) += module.o
-
-$(obj)/bootflag.c: 
-	@ln -sf ../../i386/kernel/bootflag.c $(obj)/bootflag.c
-$(obj)/cpuid.c: 
-	@ln -sf ../../i386/kernel/cpuid.c $(obj)/cpuid.c
-
-clean-files += bootflag.c cpuid.c
+obj-$(CONFIG_MODULES)		+= module.o
 
+bootflag-y			+= ../../i386/kernel/bootflag.o
+cpuid-$(CONFIG_X86_CPUID)	+= ../../i386/kernel/cpuid.o

arch/x86_64/kernel/time.c

@@ -374,10 +374,21 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id, struct pt_regs *regs)
    and could overflow, it may be better (but slower)to use an 64bit division. */
 unsigned long long sched_clock(void)
 {
-	unsigned long a;
+	unsigned long a = 0;
 
+#if 0
+	/* Don't do a HPET read here. Using TSC always is much faster
+	   and HPET may not be mapped yet when the scheduler first runs.
+           Disadvantage is a small drift between CPUs in some configurations,
+	   but that should be tolerable. */
 	if (__vxtime.mode == VXTIME_HPET)
 		return (hpet_readl(HPET_COUNTER) * vxtime.quot) >> 32;
+#endif
+
+	/* Could do CPU core sync here. Opteron can execute rdtsc speculatively,
+	   which means it is not completely exact and may not be monotonous between
+	   CPUs. But the errors should be too small to matter for scheduling
+	   purposes. */
 
 	rdtscll(a);
 	return (a * vxtime.tsc_quot) >> 32;

arch/x86_64/mm/fault.c

@@ -31,6 +31,7 @@
 #include <asm/smp.h>
 #include <asm/tlbflush.h>
 #include <asm/proto.h>
+#include <asm/kdebug.h>
 
 void bust_spinlocks(int yes)
 {

arch/x86_64/mm/pageattr.c

@@ -57,7 +57,7 @@ static void flush_kernel_map(void *address)
 		/* is this worth it? */ 
 		int i;
 		for (i = 0; i < PAGE_SIZE; i += boot_cpu_data.x86_clflush_size) 
-			asm volatile("clflush %0" :: "m" (address + i)); 
+			asm volatile("clflush (%0)" :: "r" (address + i)); 
 	} else
 		asm volatile("wbinvd":::"memory"); 
 	__flush_tlb_one(address);

arch/x86_64/oprofile/Makefile

@@ -11,24 +11,9 @@ DRIVER_OBJS = $(addprefix ../../../drivers/oprofile/, \
 	event_buffer.o oprofile_files.o \
 	oprofilefs.o oprofile_stats.o \
 	timer_int.o )
- 
-oprofile-objs := $(DRIVER_OBJS) init.o
 
-oprofile-$(CONFIG_X86_LOCAL_APIC) += nmi_int.o op_model_athlon.o
-oprofile-$(CONFIG_X86_IO_APIC)          += nmi_timer_int.o 
-INCL := $(obj)/op_counter.h $(obj)/op_x86_model.h
+OPROFILE-y := init.o
+OPROFILE-$(CONFIG_X86_LOCAL_APIC) += nmi_int.o op_model_athlon.o
+OPROFILE-$(CONFIG_X86_IO_APIC)    += nmi_timer_int.o 
 
-$(obj)/nmi_int.c: ${INCL}
-	@ln -sf ../../i386/oprofile/nmi_int.c $(obj)/nmi_int.c
-$(obj)/op_model_athlon.c: ${INCL}
-	@ln -sf ../../i386/oprofile/op_model_athlon.c $(obj)/op_model_athlon.c
-$(obj)/init.c: ${INCL}
-	@ln -sf ../../i386/oprofile/init.c $(obj)/init.c
-$(obj)/nmi_timer_int.c: ${INCL}
-	@ln -sf ../../i386/oprofile/nmi_timer_int.c $(obj)/nmi_timer_int.c
-$(obj)/op_counter.h: 
-	@ln -sf ../../i386/oprofile/op_counter.h $(obj)/op_counter.h
-$(obj)/op_x86_model.h:
-	@ln -sf ../../i386/oprofile/op_x86_model.h $(obj)/op_x86_model.h	
-clean-files += op_x86_model.h op_counter.h init.c \
-	       op_model_athlon.c nmi_int.c nmi_timer_int.c
+oprofile-y = $(DRIVER_OBJS) $(addprefix ../../i386/oprofile/, $(OPROFILE-y))