Commit 537b43ba authored by David Mosberger's avatar David Mosberger

Merge tiger.hpl.hp.com:/data1/bk/vanilla/linux-2.5

into tiger.hpl.hp.com:/data1/bk/lia64/to-linus-2.5
parents 67cba4d4 fa6ec6f7
......@@ -102,30 +102,14 @@ compressed: vmlinux
rawboot:
@$(MAKEBOOT) rawboot
#
# My boot writes directly to a specific disk partition, I doubt most
# people will want to do that without changes..
#
msb my-special-boot:
@$(MAKEBOOT) msb
bootimage:
@$(MAKEBOOT) bootimage
srmboot:
@$(MAKEBOOT) srmboot
archclean:
@$(MAKEBOOT) clean
archmrproper:
@$(MAKE) -C arch/$(ARCH)/tools mrproper
bootpfile:
@$(MAKEBOOT) bootpfile
prepare: $(TOPDIR)/include/asm-ia64/offsets.h
$(TOPDIR)/include/asm-ia64/offsets.h: include/asm include/linux/version.h \
include/config/MARKER
@$(MAKE) -C arch/$(ARCH)/tools $@
\ No newline at end of file
@$(MAKE) -C arch/$(ARCH)/tools $@
......@@ -86,6 +86,31 @@ fi
define_bool CONFIG_KCORE_ELF y # On IA-64, we always want an ELF /proc/kcore.
define_int CONFIG_FORCE_MAX_ZONEORDER 18
bool 'IA-64 Huge TLB Page Support' CONFIG_HUGETLB_PAGE
if [ "$CONFIG_HUGETLB_PAGE" = "y" ]; then
if [ "$CONFIG_MCKINLEY" = "y" ]; then
choice ' IA-64 Huge TLB Page Size' \
"4GB CONFIG_HUGETLB_PAGE_SIZE_4GB \
256MB CONFIG_HUGETLB_PAGE_SIZE_256MB \
64MB CONFIG_HUGETLB_PAGE_SIZE_64MB \
16MB CONFIG_HUGETLB_PAGE_SIZE_16MB \
4MB CONFIG_HUGETLB_PAGE_SIZE_4MB \
1MB CONFIG_HUGETLB_PAGE_SIZE_1MB \
256KB CONFIG_HUGETLB_PAGE_SIZE_256KB" 16MB
else
choice ' IA-64 Huge TLB Page Size' \
"256MB CONFIG_HUGETLB_PAGE_SIZE_256MB \
64MB CONFIG_HUGETLB_PAGE_SIZE_64MB \
16MB CONFIG_HUGETLB_PAGE_SIZE_16MB \
4MB CONFIG_HUGETLB_PAGE_SIZE_4MB \
1MB CONFIG_HUGETLB_PAGE_SIZE_1MB \
256KB CONFIG_HUGETLB_PAGE_SIZE_256KB" 16MB
fi
fi
bool 'SMP support' CONFIG_SMP
bool 'Support running of Linux/x86 binaries' CONFIG_IA32_SUPPORT
bool 'Performance monitor support' CONFIG_PERFMON
......
......@@ -39,9 +39,7 @@
#define DRIVER_NAME "SBA"
#ifndef CONFIG_IA64_HP_PROTO
#define ALLOW_IOV_BYPASS
#endif
#define ENABLE_MARK_CLEAN
/*
** The number of debug flags is a clue - this code is fragile.
......@@ -740,6 +738,9 @@ sba_map_single(struct pci_dev *dev, void *addr, size_t size, int direction)
unsigned long phys_addr = virt_to_phys(addr);
#endif
if (!sba_list)
panic("sba_map_single: no SBA found!\n");
ioc = GET_IOC(dev);
ASSERT(ioc);
......@@ -828,6 +829,9 @@ void sba_unmap_single(struct pci_dev *dev, dma_addr_t iova, size_t size,
unsigned long flags;
dma_addr_t offset;
if (!sba_list)
panic("sba_map_single: no SBA found!\n");
ioc = GET_IOC(dev);
ASSERT(ioc);
......@@ -1094,6 +1098,9 @@ int sba_map_sg(struct pci_dev *dev, struct scatterlist *sglist, int nents,
DBG_RUN_SG("%s() START %d entries, 0x%p,0x%x\n", __FUNCTION__, nents,
sba_sg_address(sglist), sba_sg_len(sglist));
if (!sba_list)
panic("sba_map_single: no SBA found!\n");
ioc = GET_IOC(dev);
ASSERT(ioc);
......@@ -1183,6 +1190,9 @@ void sba_unmap_sg(struct pci_dev *dev, struct scatterlist *sglist, int nents,
DBG_RUN_SG("%s() START %d entries, 0x%p,0x%x\n",
__FUNCTION__, nents, sba_sg_address(sglist), sba_sg_len(sglist));
if (!sba_list)
panic("sba_map_single: no SBA found!\n");
ioc = GET_IOC(dev);
ASSERT(ioc);
......@@ -1252,10 +1262,6 @@ sba_ioc_init(struct sba_device *sba_dev, struct ioc *ioc, int ioc_num)
** Firmware programs the maximum IOV space size into the imask reg
*/
iova_space_size = ~(READ_REG(ioc->ioc_hpa + IOC_IMASK) & 0xFFFFFFFFUL) + 1;
#ifdef CONFIG_IA64_HP_PROTO
if (!iova_space_size)
iova_space_size = GB(1);
#endif
/*
** iov_order is always based on a 1GB IOVA space since we want to
......@@ -1484,66 +1490,65 @@ sba_common_init(struct sba_device *sba_dev)
#ifdef CONFIG_PROC_FS
static int sba_proc_info(char *buf, char **start, off_t offset, int len)
{
struct sba_device *sba_dev = sba_list;
struct ioc *ioc = &sba_dev->ioc[0]; /* FIXME: Multi-IOC support! */
int total_pages = (int) (ioc->res_size << 3); /* 8 bits per byte */
struct sba_device *sba_dev;
struct ioc *ioc;
int total_pages;
unsigned long i = 0, avg = 0, min, max;
sprintf(buf, "%s rev %d.%d\n",
"Hewlett Packard zx1 SBA",
((sba_dev->hw_rev >> 4) & 0xF),
(sba_dev->hw_rev & 0xF)
);
sprintf(buf, "%sIO PDIR size : %d bytes (%d entries)\n",
buf,
(int) ((ioc->res_size << 3) * sizeof(u64)), /* 8 bits/byte */
total_pages);
sprintf(buf, "%sIO PDIR entries : %ld free %ld used (%d%%)\n", buf,
total_pages - ioc->used_pages, ioc->used_pages,
(int) (ioc->used_pages * 100 / total_pages));
sprintf(buf, "%sResource bitmap : %d bytes (%d pages)\n",
buf, ioc->res_size, ioc->res_size << 3); /* 8 bits per byte */
min = max = ioc->avg_search[0];
for (i = 0; i < SBA_SEARCH_SAMPLE; i++) {
avg += ioc->avg_search[i];
if (ioc->avg_search[i] > max) max = ioc->avg_search[i];
if (ioc->avg_search[i] < min) min = ioc->avg_search[i];
}
avg /= SBA_SEARCH_SAMPLE;
sprintf(buf, "%s Bitmap search : %ld/%ld/%ld (min/avg/max CPU Cycles)\n",
buf, min, avg, max);
for (sba_dev = sba_list; sba_dev; sba_dev = sba_dev->next) {
ioc = &sba_dev->ioc[0]; /* FIXME: Multi-IOC support! */
total_pages = (int) (ioc->res_size << 3); /* 8 bits per byte */
sprintf(buf, "%s rev %d.%d\n", "Hewlett Packard zx1 SBA",
((sba_dev->hw_rev >> 4) & 0xF), (sba_dev->hw_rev & 0xF));
sprintf(buf, "%sIO PDIR size : %d bytes (%d entries)\n", buf,
(int) ((ioc->res_size << 3) * sizeof(u64)), /* 8 bits/byte */ total_pages);
sprintf(buf, "%sIO PDIR entries : %ld free %ld used (%d%%)\n", buf,
total_pages - ioc->used_pages, ioc->used_pages,
(int) (ioc->used_pages * 100 / total_pages));
sprintf(buf, "%sResource bitmap : %d bytes (%d pages)\n",
buf, ioc->res_size, ioc->res_size << 3); /* 8 bits per byte */
sprintf(buf, "%spci_map_single(): %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->msingle_calls, ioc->msingle_pages,
(int) ((ioc->msingle_pages * 1000)/ioc->msingle_calls));
min = max = ioc->avg_search[0];
for (i = 0; i < SBA_SEARCH_SAMPLE; i++) {
avg += ioc->avg_search[i];
if (ioc->avg_search[i] > max) max = ioc->avg_search[i];
if (ioc->avg_search[i] < min) min = ioc->avg_search[i];
}
avg /= SBA_SEARCH_SAMPLE;
sprintf(buf, "%s Bitmap search : %ld/%ld/%ld (min/avg/max CPU Cycles)\n",
buf, min, avg, max);
sprintf(buf, "%spci_map_single(): %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->msingle_calls, ioc->msingle_pages,
(int) ((ioc->msingle_pages * 1000)/ioc->msingle_calls));
#ifdef ALLOW_IOV_BYPASS
sprintf(buf, "%spci_map_single(): %12ld bypasses\n",
buf, ioc->msingle_bypass);
sprintf(buf, "%spci_map_single(): %12ld bypasses\n",
buf, ioc->msingle_bypass);
#endif
sprintf(buf, "%spci_unmap_single: %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->usingle_calls, ioc->usingle_pages,
(int) ((ioc->usingle_pages * 1000)/ioc->usingle_calls));
sprintf(buf, "%spci_unmap_single: %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->usingle_calls, ioc->usingle_pages,
(int) ((ioc->usingle_pages * 1000)/ioc->usingle_calls));
#ifdef ALLOW_IOV_BYPASS
sprintf(buf, "%spci_unmap_single: %12ld bypasses\n",
buf, ioc->usingle_bypass);
sprintf(buf, "%spci_unmap_single: %12ld bypasses\n",
buf, ioc->usingle_bypass);
#endif
sprintf(buf, "%spci_map_sg() : %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->msg_calls, ioc->msg_pages,
(int) ((ioc->msg_pages * 1000)/ioc->msg_calls));
sprintf(buf, "%spci_map_sg() : %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->msg_calls, ioc->msg_pages,
(int) ((ioc->msg_pages * 1000)/ioc->msg_calls));
#ifdef ALLOW_IOV_BYPASS
sprintf(buf, "%spci_map_sg() : %12ld bypasses\n",
buf, ioc->msg_bypass);
sprintf(buf, "%spci_map_sg() : %12ld bypasses\n",
buf, ioc->msg_bypass);
#endif
sprintf(buf, "%spci_unmap_sg() : %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->usg_calls, ioc->usg_pages,
(int) ((ioc->usg_pages * 1000)/ioc->usg_calls));
sprintf(buf, "%spci_unmap_sg() : %12ld calls %12ld pages (avg %d/1000)\n",
buf, ioc->usg_calls, ioc->usg_pages,
(int) ((ioc->usg_pages * 1000)/ioc->usg_calls));
}
return strlen(buf);
}
......@@ -1551,9 +1556,13 @@ static int
sba_resource_map(char *buf, char **start, off_t offset, int len)
{
struct ioc *ioc = sba_list->ioc; /* FIXME: Multi-IOC support! */
unsigned int *res_ptr = (unsigned int *)ioc->res_map;
unsigned int *res_ptr;
int i;
if (!ioc)
return 0;
res_ptr = (unsigned int *)ioc->res_map;
buf[0] = '\0';
for(i = 0; i < (ioc->res_size / sizeof(unsigned int)); ++i, ++res_ptr) {
if ((i & 7) == 0)
......@@ -1593,13 +1602,11 @@ void __init sba_init(void)
}
func_id = READ_REG(hpa + SBA_FUNC_ID);
if (func_id == ZX1_FUNC_ID_VALUE) {
strcpy(sba_rev, "zx1");
func_offset = zx1_func_offsets;
} else {
if (func_id != ZX1_FUNC_ID_VALUE)
return;
}
strcpy(sba_rev, "zx1");
func_offset = zx1_func_offsets;
/* Read HW Rev First */
hw_rev = READ_REG(hpa + SBA_FCLASS) & 0xFFUL;
......@@ -1625,10 +1632,8 @@ void __init sba_init(void)
device->slot_name, hpa);
if ((hw_rev & 0xFF) < 0x20) {
printk(KERN_INFO "%s WARNING rev 2.0 or greater will be required for IO MMU support in the future\n", DRIVER_NAME);
#ifndef CONFIG_IA64_HP_PROTO
panic("%s: CONFIG_IA64_HP_PROTO MUST be enabled to support SBA rev less than 2.0", DRIVER_NAME);
#endif
printk("%s: SBA rev less than 2.0 not supported", DRIVER_NAME);
return;
}
sba_dev = kmalloc(sizeof(struct sba_device), GFP_KERNEL);
......
......@@ -25,68 +25,84 @@ extern acpi_status acpi_evaluate_integer (acpi_handle, acpi_string, acpi_object_
static int hpzx1_devices;
struct fake_pci_dev {
struct fake_pci_dev *next;
struct pci_dev *pci_dev;
unsigned long csr_base;
unsigned long csr_size;
unsigned long mapped_csrs; // ioremapped
int sizing; // in middle of BAR sizing operation?
};
} *fake_pci_dev_list;
static struct pci_ops *orig_pci_ops;
#define HP_CFG_RD(sz, bits, name) \
static int hp_cfg_read##sz (struct pci_dev *dev, int where, u##bits *value) \
{ \
struct fake_pci_dev *fake_dev; \
if (!(fake_dev = (struct fake_pci_dev *) dev->sysdata)) \
return orig_pci_ops->name(dev, where, value); \
\
if (where == PCI_BASE_ADDRESS_0) { \
if (fake_dev->sizing) \
*value = ~(fake_dev->csr_size - 1); \
else \
*value = (fake_dev->csr_base & \
PCI_BASE_ADDRESS_MEM_MASK) | \
PCI_BASE_ADDRESS_SPACE_MEMORY; \
fake_dev->sizing = 0; \
return PCIBIOS_SUCCESSFUL; \
} \
*value = read##sz(fake_dev->mapped_csrs + where); \
if (where == PCI_COMMAND) \
*value |= PCI_COMMAND_MEMORY; /* SBA omits this */ \
return PCIBIOS_SUCCESSFUL; \
struct fake_pci_dev *
lookup_fake_dev (struct pci_bus *bus, unsigned int devfn)
{
struct fake_pci_dev *fake_dev;
for (fake_dev = fake_pci_dev_list; fake_dev; fake_dev = fake_dev->next)
if (fake_dev->pci_dev->bus == bus && fake_dev->pci_dev->devfn == devfn)
return fake_dev;
return NULL;
}
#define HP_CFG_WR(sz, bits, name) \
static int hp_cfg_write##sz (struct pci_dev *dev, int where, u##bits value) \
{ \
struct fake_pci_dev *fake_dev; \
\
if (!(fake_dev = (struct fake_pci_dev *) dev->sysdata)) \
return orig_pci_ops->name(dev, where, value); \
\
if (where == PCI_BASE_ADDRESS_0) { \
if (value == (u##bits) ~0) \
fake_dev->sizing = 1; \
return PCIBIOS_SUCCESSFUL; \
} else \
write##sz(value, fake_dev->mapped_csrs + where); \
return PCIBIOS_SUCCESSFUL; \
static int
hp_cfg_read (struct pci_bus *bus, unsigned int devfn, int where, int size, u32 *value)
{
struct fake_pci_dev *fake_dev = lookup_fake_dev(bus, devfn);
if (!fake_dev)
return (*orig_pci_ops->read)(bus, devfn, where, size, value);
if (where == PCI_BASE_ADDRESS_0) {
if (fake_dev->sizing)
*value = ~(fake_dev->csr_size - 1);
else
*value = ((fake_dev->csr_base & PCI_BASE_ADDRESS_MEM_MASK)
| PCI_BASE_ADDRESS_SPACE_MEMORY);
fake_dev->sizing = 0;
return PCIBIOS_SUCCESSFUL;
}
switch (size) {
case 1: *value = readb(fake_dev->mapped_csrs + where); break;
case 2: *value = readw(fake_dev->mapped_csrs + where); break;
case 4: *value = readl(fake_dev->mapped_csrs + where); break;
default:
printk(KERN_WARNING"hp_cfg_read: bad size = %d bytes", size);
break;
}
if (where == PCI_COMMAND)
*value |= PCI_COMMAND_MEMORY; /* SBA omits this */
return PCIBIOS_SUCCESSFUL;
}
HP_CFG_RD(b, 8, read_byte)
HP_CFG_RD(w, 16, read_word)
HP_CFG_RD(l, 32, read_dword)
HP_CFG_WR(b, 8, write_byte)
HP_CFG_WR(w, 16, write_word)
HP_CFG_WR(l, 32, write_dword)
static int
hp_cfg_write (struct pci_bus *bus, unsigned int devfn, int where, int size, u32 value)
{
struct fake_pci_dev *fake_dev = lookup_fake_dev(bus, devfn);
if (!fake_dev)
return (*orig_pci_ops->write)(bus, devfn, where, size, value);
if (where == PCI_BASE_ADDRESS_0) {
if (value == ((1UL << 8*size) - 1))
fake_dev->sizing = 1;
return PCIBIOS_SUCCESSFUL;
}
switch (size) {
case 1: writeb(value, fake_dev->mapped_csrs + where); break;
case 2: writew(value, fake_dev->mapped_csrs + where); break;
case 4: writel(value, fake_dev->mapped_csrs + where); break;
default:
printk(KERN_WARNING"hp_cfg_write: bad size = %d bytes", size);
break;
}
return PCIBIOS_SUCCESSFUL;
}
static struct pci_ops hp_pci_conf = {
hp_cfg_readb,
hp_cfg_readw,
hp_cfg_readl,
hp_cfg_writeb,
hp_cfg_writew,
hp_cfg_writel,
.read = hp_cfg_read,
.write = hp_cfg_write
};
static void
......@@ -143,6 +159,10 @@ hpzx1_fake_pci_dev(char *name, unsigned int busnum, unsigned long addr, unsigned
bus->ops = &hp_pci_conf; // replace pci ops for this bus
fake->pci_dev = dev;
fake->next = fake_pci_dev_list;
fake_pci_dev_list = fake;
memset(dev, 0, sizeof(*dev));
dev->bus = bus;
dev->sysdata = fake;
......@@ -309,40 +329,8 @@ hpzx1_acpi_dev_init(void)
* HWP0003: AGP LBA device
*/
acpi_get_devices("HWP0001", hpzx1_sba_probe, "HWP0001", NULL);
#ifdef CONFIG_IA64_HP_PROTO
if (hpzx1_devices) {
#endif
acpi_get_devices("HWP0002", hpzx1_lba_probe, "HWP0002 PCI LBA", NULL);
acpi_get_devices("HWP0003", hpzx1_lba_probe, "HWP0003 AGP LBA", NULL);
#ifdef CONFIG_IA64_HP_PROTO
}
#define ZX1_FUNC_ID_VALUE (PCI_DEVICE_ID_HP_ZX1_SBA << 16) | PCI_VENDOR_ID_HP
/*
* Early protos don't have bridges in the ACPI namespace, so
* if we didn't find anything, add the things we know are
* there.
*/
if (hpzx1_devices == 0) {
u64 hpa, csr_base;
csr_base = 0xfed00000UL;
hpa = (u64) ioremap(csr_base, 0x2000);
if (__raw_readl(hpa) == ZX1_FUNC_ID_VALUE) {
hpzx1_fake_pci_dev("HWP0001 SBA", 0, csr_base, 0x1000);
hpzx1_fake_pci_dev("HWP0001 IOC", 0, csr_base + 0x1000,
0x1000);
csr_base = 0xfed24000UL;
iounmap(hpa);
hpa = (u64) ioremap(csr_base, 0x1000);
hpzx1_fake_pci_dev("HWP0003 AGP LBA", 0x40, csr_base,
0x1000);
}
iounmap(hpa);
}
#endif
}
extern void sba_init(void);
......
......@@ -2111,8 +2111,8 @@ struct shm_info32 {
};
struct ipc_kludge {
struct msgbuf *msgp;
long msgtyp;
u32 msgp;
s32 msgtyp;
};
#define SEMOP 1
......
......@@ -33,6 +33,15 @@
#define EFI_DEBUG 0
#ifdef CONFIG_HUGETLB_PAGE
/* By default at total of 512MB is reserved huge pages. */
#define HTLBZONE_SIZE_DEFAULT 0x20000000
unsigned long htlbzone_pages = (HTLBZONE_SIZE_DEFAULT >> HPAGE_SHIFT);
#endif
extern efi_status_t efi_call_phys (void *, ...);
struct efi efi;
......@@ -399,6 +408,25 @@ efi_init (void)
++cp;
}
}
#ifdef CONFIG_HUGETLB_PAGE
/* Just duplicating the above algo for lpzone start */
for (cp = saved_command_line; *cp; ) {
if (memcmp(cp, "lpmem=", 6) == 0) {
cp += 6;
htlbzone_pages = memparse(cp, &end);
htlbzone_pages = (htlbzone_pages >> HPAGE_SHIFT);
if (end != cp)
break;
cp = end;
} else {
while (*cp != ' ' && *cp)
++cp;
while (*cp == ' ')
++cp;
}
}
printk("Total HugeTLB_Page memory pages requested 0x%lx \n", htlbzone_pages);
#endif
if (mem_limit != ~0UL)
printk("Ignoring memory above %luMB\n", mem_limit >> 20);
......
......@@ -90,15 +90,23 @@ ENTRY(ia64_execve)
br.ret.sptk.many rp
END(ia64_execve)
/*
* sys_clone2(u64 flags, u64 ustack_base, u64 ustack_size, u64 user_tid, u64 tls)
*/
GLOBAL_ENTRY(sys_clone2)
.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(2)
alloc r16=ar.pfs,3,2,4,0
alloc r16=ar.pfs,5,2,5,0
DO_SAVE_SWITCH_STACK
adds r2=PT(R16)+IA64_SWITCH_STACK_SIZE+16,sp
mov loc0=rp
mov loc1=r16 // save ar.pfs across do_fork
.body
mov out1=in1
mov out3=in2
tbit.nz p6,p0=in0,CLONE_SETTLS_BIT
mov out4=in3 // valid only w/CLONE_SETTID and/or CLONE_CLEARTID
;;
(p6) st8 [r2]=in4 // store TLS in r13 (tp)
adds out2=IA64_SWITCH_STACK_SIZE+16,sp // out2 = &regs
dep out0=0,in0,CLONE_IDLETASK_BIT,1 // out0 = clone_flags & ~CLONE_IDLETASK
br.call.sptk.many rp=do_fork
......@@ -115,15 +123,24 @@ GLOBAL_ENTRY(sys_clone2)
br.ret.sptk.many rp
END(sys_clone2)
/*
* sys_clone(u64 flags, u64 ustack_base, u64 user_tid, u64 tls)
* Deprecated. Use sys_clone2() instead.
*/
GLOBAL_ENTRY(sys_clone)
.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(2)
alloc r16=ar.pfs,2,2,4,0
alloc r16=ar.pfs,4,2,5,0
DO_SAVE_SWITCH_STACK
adds r2=PT(R16)+IA64_SWITCH_STACK_SIZE+16,sp
mov loc0=rp
mov loc1=r16 // save ar.pfs across do_fork
.body
mov out1=in1
mov out3=16 // stacksize (compensates for 16-byte scratch area)
tbit.nz p6,p0=in0,CLONE_SETTLS_BIT
mov out4=in2 // out4 = user_tid (optional)
;;
(p6) st8 [r2]=in3 // store TLS in r13 (tp)
adds out2=IA64_SWITCH_STACK_SIZE+16,sp // out2 = &regs
dep out0=0,in0,CLONE_IDLETASK_BIT,1 // out0 = clone_flags & ~CLONE_IDLETASK
br.call.sptk.many rp=do_fork
......@@ -521,7 +538,7 @@ GLOBAL_ENTRY(ia64_ret_from_clone)
#if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT)
/*
* We need to call schedule_tail() to complete the scheduling process.
* Called by ia64_switch_to after do_fork()->copy_thread(). r8 contains the
* Called by ia64_switch_to() after do_fork()->copy_thread(). r8 contains the
* address of the previously executing task.
*/
br.call.sptk.many rp=ia64_invoke_schedule_tail
......@@ -872,7 +889,7 @@ END(invoke_schedule)
#endif /* __GNUC__ < 3 */
/*
* Setup stack and call ia64_do_signal. Note that pSys and pNonSys need to
* Setup stack and call do_notify_resume_user(). Note that pSys and pNonSys need to
* be set up by the caller. We declare 8 input registers so the system call
* args get preserved, in case we need to restart a system call.
*/
......@@ -900,7 +917,7 @@ ENTRY(notify_resume_user)
mov ar.unat=r9
mov ar.pfs=loc1
br.ret.sptk.many rp
END(do_notify_resume_user)
END(notify_resume_user)
GLOBAL_ENTRY(sys_rt_sigsuspend)
.prologue ASM_UNW_PRLG_RP|ASM_UNW_PRLG_PFS, ASM_UNW_PRLG_GRSAVE(8)
......@@ -1224,10 +1241,10 @@ sys_call_table:
data8 sys_futex // 1230
data8 sys_sched_setaffinity
data8 sys_sched_getaffinity
data8 ia64_ni_syscall
data8 ia64_ni_syscall
data8 ia64_ni_syscall // 1235
data8 ia64_ni_syscall
data8 sys_security
data8 sys_alloc_hugepages
data8 sys_free_hugepages // 1235
data8 sys_exit_group
data8 ia64_ni_syscall
data8 sys_io_setup
data8 sys_io_destroy
......
......@@ -2,7 +2,7 @@
* This file contains the code that gets mapped at the upper end of each task's text
* region. For now, it contains the signal trampoline code only.
*
* Copyright (C) 1999-2001 Hewlett-Packard Co
* Copyright (C) 1999-2002 Hewlett-Packard Co
* David Mosberger-Tang <davidm@hpl.hp.com>
*/
......@@ -135,7 +135,7 @@ back_from_setup_rbs:
;;
ld8 r8=[base0] // restore (perhaps modified) CFM0, EC0, and CPL0
cmp.ne p8,p0=r14,r15 // do we need to restore the rbs?
(p8) br.cond.spnt restore_rbs // yup -> (clobbers r14 and r16)
(p8) br.cond.spnt restore_rbs // yup -> (clobbers r14-r18, f6 & f7)
;;
back_from_restore_rbs:
adds base0=(FR6_OFF+SIGCONTEXT_OFF),sp
......@@ -189,20 +189,69 @@ setup_rbs:
.spillsp ar.rnat, RNAT_OFF+SIGCONTEXT_OFF
.body
restore_rbs:
// On input:
// r14 = bsp1 (bsp at the time of return from signal handler)
// r15 = bsp0 (bsp at the time the signal occurred)
//
// Here, we need to calculate bspstore0, the value that ar.bspstore needs
// to be set to, based on bsp0 and the size of the dirty partition on
// the alternate stack (sc_loadrs >> 16). This can be done with the
// following algorithm:
//
// bspstore0 = rse_skip_regs(bsp0, -rse_num_regs(bsp1 - (loadrs >> 19), bsp1));
//
// This is what the code below does.
//
alloc r2=ar.pfs,0,0,0,0 // alloc null frame
adds r16=(LOADRS_OFF+SIGCONTEXT_OFF),sp
adds r18=(RNAT_OFF+SIGCONTEXT_OFF),sp
;;
ld8 r14=[r16]
adds r16=(RNAT_OFF+SIGCONTEXT_OFF),sp
ld8 r17=[r16]
ld8 r16=[r18] // get new rnat
extr.u r18=r15,3,6 // r18 <- rse_slot_num(bsp0)
;;
mov ar.rsc=r14 // put RSE into enforced lazy mode
ld8 r14=[r16] // get new rnat
mov ar.rsc=r17 // put RSE into enforced lazy mode
shr.u r17=r17,16
;;
loadrs // restore dirty partition
sub r14=r14,r17 // r14 (bspstore1) <- bsp1 - (sc_loadrs >> 16)
shr.u r17=r17,3 // r17 <- (sc_loadrs >> 19)
;;
loadrs // restore dirty partition
extr.u r14=r14,3,6 // r14 <- rse_slot_num(bspstore1)
;;
add r14=r14,r17 // r14 <- rse_slot_num(bspstore1) + (sc_loadrs >> 19)
;;
shr.u r14=r14,6 // r14 <- (rse_slot_num(bspstore1) + (sc_loadrs >> 19))/0x40
;;
sub r14=r14,r17 // r14 <- -rse_num_regs(bspstore1, bsp1)
movl r17=0x8208208208208209
;;
add r18=r18,r14 // r18 (delta) <- rse_slot_num(bsp0) - rse_num_regs(bspstore1,bsp1)
setf.sig f7=r17
cmp.lt p7,p0=r14,r0 // p7 <- (r14 < 0)?
;;
(p7) adds r18=-62,r18 // delta -= 62
;;
setf.sig f6=r18
;;
xmpy.h f6=f6,f7
;;
getf.sig r17=f6
;;
add r17=r17,r18
shr r18=r18,63
;;
shr r17=r17,5
;;
sub r17=r17,r18 // r17 = delta/63
;;
add r17=r14,r17 // r17 <- delta/63 - rse_num_regs(bspstore1, bsp1)
;;
shladd r15=r17,3,r15 // r15 <- bsp0 + 8*(delta/63 - rse_num_regs(bspstore1, bsp1))
;;
mov ar.bspstore=r15 // switch back to old register backing store area
;;
mov ar.rnat=r14 // restore RNaT
mov ar.rnat=r16 // restore RNaT
mov ar.rsc=0xf // (will be restored later on from sc_ar_rsc)
// invala not necessary as that will happen when returning to user-mode
br.cond.sptk back_from_restore_rbs
......
This diff is collapsed.
......@@ -127,6 +127,8 @@ EXPORT_SYMBOL(ia64_pal_call_phys_stacked);
EXPORT_SYMBOL(ia64_pal_call_phys_static);
EXPORT_SYMBOL(ia64_pal_call_stacked);
EXPORT_SYMBOL(ia64_pal_call_static);
EXPORT_SYMBOL(ia64_load_scratch_fpregs);
EXPORT_SYMBOL(ia64_save_scratch_fpregs);
extern struct efi efi;
EXPORT_SYMBOL(efi);
......
......@@ -16,7 +16,7 @@
static struct fs_struct init_fs = INIT_FS;
static struct files_struct init_files = INIT_FILES;
static struct signal_struct init_signals = INIT_SIGNALS;
static struct signal_struct init_signals = INIT_SIGNALS(init_signals);
struct mm_struct init_mm = INIT_MM(init_mm);
/*
......
......@@ -403,8 +403,8 @@ unsigned int do_IRQ(unsigned long irq, struct pt_regs *regs)
break;
desc->status &= ~IRQ_PENDING;
}
out:
desc->status &= ~IRQ_INPROGRESS;
out:
/*
* The ->end() handler has to deal with interrupts which got
* disabled while the handler was running.
......@@ -788,7 +788,7 @@ int setup_irq(unsigned int irq, struct irqaction * new)
if (!shared) {
desc->depth = 0;
desc->status &= ~(IRQ_DISABLED | IRQ_AUTODETECT | IRQ_WAITING);
desc->status &= ~(IRQ_DISABLED | IRQ_AUTODETECT | IRQ_WAITING | IRQ_INPROGRESS);
desc->handler->startup(irq);
}
spin_unlock_irqrestore(&desc->lock,flags);
......
......@@ -70,24 +70,6 @@
mov r19=n;; /* prepare to save predicates */ \
br.sptk.many dispatch_to_fault_handler
/*
* As we don't (hopefully) use the space available, we need to fill it with
* nops. the parameter may be used for debugging and is representing the entry
* number
*/
#define BREAK_BUNDLE(a) break.m (a); \
break.i (a); \
break.i (a)
/*
* 4 breaks bundles all together
*/
#define BREAK_BUNDLE4(a); BREAK_BUNDLE(a); BREAK_BUNDLE(a); BREAK_BUNDLE(a); BREAK_BUNDLE(a)
/*
* 8 bundles all together (too lazy to use only 4 at a time !)
*/
#define BREAK_BUNDLE8(a); BREAK_BUNDLE4(a); BREAK_BUNDLE4(a)
.section .text.ivt,"ax"
.align 32768 // align on 32KB boundary
......@@ -115,6 +97,10 @@ ENTRY(vhpt_miss)
* - the faulting virtual address has no L1, L2, or L3 mapping
*/
mov r16=cr.ifa // get address that caused the TLB miss
#ifdef CONFIG_HUGETLB_PAGE
movl r18=PAGE_SHIFT
mov r25=cr.itir
#endif
;;
rsm psr.dt // use physical addressing for data
mov r31=pr // save the predicate registers
......@@ -122,8 +108,18 @@ ENTRY(vhpt_miss)
shl r21=r16,3 // shift bit 60 into sign bit
shr.u r17=r16,61 // get the region number into r17
;;
shr r22=r21,3
#ifdef CONFIG_HUGETLB_PAGE
extr.u r26=r25,2,6
;;
cmp.eq p8,p0=HPAGE_SHIFT,r26
;;
(p8) dep r25=r18,r25,2,6
(p8) shr r22=r22,HPAGE_SHIFT-PAGE_SHIFT
;;
#endif
cmp.eq p6,p7=5,r17 // is IFA pointing into to region 5?
shr.u r18=r16,PGDIR_SHIFT // get bits 33-63 of the faulting address
shr.u r18=r22,PGDIR_SHIFT // get bits 33-63 of the faulting address
;;
(p7) dep r17=r17,r19,(PAGE_SHIFT-3),3 // put region number bits in place
srlz.d // ensure "rsm psr.dt" has taken effect
......@@ -134,7 +130,7 @@ ENTRY(vhpt_miss)
(p6) dep r17=r18,r19,3,(PAGE_SHIFT-3) // r17=PTA + IFA(33,42)*8
(p7) dep r17=r18,r17,3,(PAGE_SHIFT-6) // r17=PTA + (((IFA(61,63) << 7) | IFA(33,39))*8)
cmp.eq p7,p6=0,r21 // unused address bits all zeroes?
shr.u r18=r16,PMD_SHIFT // shift L2 index into position
shr.u r18=r22,PMD_SHIFT // shift L2 index into position
;;
ld8 r17=[r17] // fetch the L1 entry (may be 0)
;;
......@@ -142,7 +138,7 @@ ENTRY(vhpt_miss)
dep r17=r18,r17,3,(PAGE_SHIFT-3) // compute address of L2 page table entry
;;
(p7) ld8 r20=[r17] // fetch the L2 entry (may be 0)
shr.u r19=r16,PAGE_SHIFT // shift L3 index into position
shr.u r19=r22,PAGE_SHIFT // shift L3 index into position
;;
(p7) cmp.eq.or.andcm p6,p7=r20,r0 // was L2 entry NULL?
dep r21=r19,r20,3,(PAGE_SHIFT-3) // compute address of L3 page table entry
......@@ -161,6 +157,10 @@ ENTRY(vhpt_miss)
(p6) br.cond.spnt.many page_fault // handle bad address/page not present (page fault)
mov cr.ifa=r22
#ifdef CONFIG_HUGETLB_PAGE
(p8) mov cr.itir=r25 // change to default page-size for VHPT
#endif
/*
* Now compute and insert the TLB entry for the virtual page table. We never
* execute in a page table page so there is no need to set the exception deferral
......
......@@ -245,3 +245,48 @@ GLOBAL_ENTRY(ia64_pal_call_phys_stacked)
br.ret.sptk.many b0
END(ia64_pal_call_phys_stacked)
/*
* Save scratch fp scratch regs which aren't saved in pt_regs already (fp10-fp15).
*
* NOTE: We need to do this since firmware (SAL and PAL) may use any of the scratch
* regs fp-low partition.
*
* Inputs:
* in0 Address of stack storage for fp regs
*/
GLOBAL_ENTRY(ia64_save_scratch_fpregs)
alloc r3=ar.pfs,1,0,0,0
add r2=16,in0
;;
stf.spill [in0] = f10,32
stf.spill [r2] = f11,32
;;
stf.spill [in0] = f12,32
stf.spill [r2] = f13,32
;;
stf.spill [in0] = f14,32
stf.spill [r2] = f15,32
br.ret.sptk.many rp
END(ia64_save_scratch_fpregs)
/*
* Load scratch fp scratch regs (fp10-fp15)
*
* Inputs:
* in0 Address of stack storage for fp regs
*/
GLOBAL_ENTRY(ia64_load_scratch_fpregs)
alloc r3=ar.pfs,1,0,0,0
add r2=16,in0
;;
ldf.fill f10 = [in0],32
ldf.fill f11 = [r2],32
;;
ldf.fill f12 = [in0],32
ldf.fill f13 = [r2],32
;;
ldf.fill f14 = [in0],32
ldf.fill f15 = [r2],32
br.ret.sptk.many rp
END(ia64_load_scratch_fpregs)
......@@ -2,6 +2,11 @@
* pci.c - Low-Level PCI Access in IA-64
*
* Derived from bios32.c of i386 tree.
*
* Copyright (C) 2002 Hewlett-Packard Co
* David Mosberger-Tang <davidm@hpl.hp.com>
*
* Note: Above list of copyright holders is incomplete...
*/
#include <linux/config.h>
......@@ -85,15 +90,15 @@ __pci_sal_write (int seg, int bus, int dev, int fn, int reg, int len, u32 value)
static int
pci_sal_read (struct pci_bus *bus, unsigned int devfn, int where, int size, u32 *value)
{
return __pci_sal_read(0, bus->number, PCI_SLOT(devfn),
PCI_FUNC(devfn), where, size, value);
return __pci_sal_read(0, bus->number, PCI_SLOT(devfn), PCI_FUNC(devfn),
where, size, value);
}
static int
pci_sal_write (struct pci_bus *bus, unsigned int devfn, int where, int size, u32 value)
{
return __pci_sal_write(0, bus->number, PCI_SLOT(devfn),
PCI_FUNC(devfn), where, size, value);
return __pci_sal_write(0, bus->number, PCI_SLOT(devfn), PCI_FUNC(devfn),
where, size, value);
}
struct pci_ops pci_sal_ops = {
......@@ -202,8 +207,8 @@ pcibios_fixup_pbus_ranges (struct pci_bus * bus, struct pbus_set_ranges_data * r
{
}
int
pcibios_enable_device (struct pci_dev *dev)
static inline int
pcibios_enable_resources (struct pci_dev *dev, int mask)
{
u16 cmd, old_cmd;
int idx;
......@@ -215,6 +220,10 @@ pcibios_enable_device (struct pci_dev *dev)
pci_read_config_word(dev, PCI_COMMAND, &cmd);
old_cmd = cmd;
for (idx=0; idx<6; idx++) {
/* Only set up the desired resources. */
if (!(mask & (1 << idx)))
continue;
r = &dev->resource[idx];
if (!r->start && r->end) {
printk(KERN_ERR
......@@ -233,9 +242,19 @@ pcibios_enable_device (struct pci_dev *dev)
printk("PCI: Enabling device %s (%04x -> %04x)\n", dev->slot_name, old_cmd, cmd);
pci_write_config_word(dev, PCI_COMMAND, cmd);
}
return 0;
}
printk(KERN_INFO "PCI: Found IRQ %d for device %s\n", dev->irq, dev->slot_name);
int
pcibios_enable_device (struct pci_dev *dev, int mask)
{
int ret;
ret = pcibios_enable_resources(dev, mask);
if (ret < 0)
return ret;
printk(KERN_INFO "PCI: Found IRQ %d for device %s\n", dev->irq, dev->slot_name);
return 0;
}
......
This diff is collapsed.
......@@ -80,6 +80,12 @@ show_stack (struct task_struct *task)
}
}
void
dump_stack (void)
{
show_stack(NULL);
}
void
show_regs (struct pt_regs *regs)
{
......@@ -248,18 +254,15 @@ ia64_load_extra (struct task_struct *task)
* | | <-- sp (lowest addr)
* +---------------------+
*
* Note: if we get called through kernel_thread() then the memory
* above "(highest addr)" is valid kernel stack memory that needs to
* be copied as well.
* Note: if we get called through kernel_thread() then the memory above "(highest addr)"
* is valid kernel stack memory that needs to be copied as well.
*
* Observe that we copy the unat values that are in pt_regs and
* switch_stack. Spilling an integer to address X causes bit N in
* ar.unat to be set to the NaT bit of the register, with N=(X &
* 0x1ff)/8. Thus, copying the unat value preserves the NaT bits ONLY
* if the pt_regs structure in the parent is congruent to that of the
* child, modulo 512. Since the stack is page aligned and the page
* size is at least 4KB, this is always the case, so there is nothing
* to worry about.
* Observe that we copy the unat values that are in pt_regs and switch_stack. Spilling an
* integer to address X causes bit N in ar.unat to be set to the NaT bit of the register,
* with N=(X & 0x1ff)/8. Thus, copying the unat value preserves the NaT bits ONLY if the
* pt_regs structure in the parent is congruent to that of the child, modulo 512. Since
* the stack is page aligned and the page size is at least 4KB, this is always the case,
* so there is nothing to worry about.
*/
int
copy_thread (int nr, unsigned long clone_flags,
......@@ -300,6 +303,8 @@ copy_thread (int nr, unsigned long clone_flags,
memcpy((void *) child_rbs, (void *) rbs, rbs_size);
if (user_mode(child_ptregs)) {
if (clone_flags & CLONE_SETTLS)
child_ptregs->r13 = regs->r16; /* see sys_clone2() in entry.S */
if (user_stack_base) {
child_ptregs->r12 = user_stack_base + user_stack_size - 16;
child_ptregs->ar_bspstore = user_stack_base;
......
......@@ -474,7 +474,7 @@ threads_sync_user_rbs (struct task_struct *child, unsigned long child_urbs_end,
{
struct switch_stack *sw;
unsigned long urbs_end;
struct task_struct *p;
struct task_struct *g, *p;
struct mm_struct *mm;
struct pt_regs *pt;
long multi_threaded;
......@@ -495,7 +495,7 @@ threads_sync_user_rbs (struct task_struct *child, unsigned long child_urbs_end,
} else {
read_lock(&tasklist_lock);
{
for_each_task(p) {
do_each_thread(g, p) {
if (p->mm == mm && p->state != TASK_RUNNING) {
sw = (struct switch_stack *) (p->thread.ksp + 16);
pt = ia64_task_regs(p);
......@@ -504,7 +504,7 @@ threads_sync_user_rbs (struct task_struct *child, unsigned long child_urbs_end,
if (make_writable)
user_flushrs(p, pt);
}
}
} while_each_thread(g, p);
}
read_unlock(&tasklist_lock);
}
......
......@@ -15,8 +15,8 @@
* test if they need to do any extra work (up needs to do something
* only if count was negative before the increment operation.
*
* "sleepers" and the contention routine ordering is protected by the
* semaphore spinlock.
* "sleeping" and the contention routine ordering is protected
* by the spinlock in the semaphore's waitqueue head.
*
* Note that these functions are only called when there is contention
* on the lock, and as such all this is the "non-critical" part of the
......@@ -44,40 +44,42 @@ __up (struct semaphore *sem)
wake_up(&sem->wait);
}
static spinlock_t semaphore_lock = SPIN_LOCK_UNLOCKED;
void
__down (struct semaphore *sem)
{
struct task_struct *tsk = current;
DECLARE_WAITQUEUE(wait, tsk);
unsigned long flags;
tsk->state = TASK_UNINTERRUPTIBLE;
add_wait_queue_exclusive(&sem->wait, &wait);
spin_lock_irqsave(&sem->wait.lock, flags);
add_wait_queue_exclusive_locked(&sem->wait, &wait);
spin_lock_irq(&semaphore_lock);
sem->sleepers++;
for (;;) {
int sleepers = sem->sleepers;
/*
* Add "everybody else" into it. They aren't
* playing, because we own the spinlock.
* playing, because we own the spinlock in
* the wait_queue_head.
*/
if (!atomic_add_negative(sleepers - 1, &sem->count)) {
sem->sleepers = 0;
break;
}
sem->sleepers = 1; /* us - see -1 above */
spin_unlock_irq(&semaphore_lock);
spin_unlock_irqrestore(&sem->wait.lock, flags);
schedule();
spin_lock_irqsave(&sem->wait.lock, flags);
tsk->state = TASK_UNINTERRUPTIBLE;
spin_lock_irq(&semaphore_lock);
}
spin_unlock_irq(&semaphore_lock);
remove_wait_queue(&sem->wait, &wait);
remove_wait_queue_locked(&sem->wait, &wait);
wake_up_locked(&sem->wait);
spin_unlock_irqrestore(&sem->wait.lock, flags);
tsk->state = TASK_RUNNING;
wake_up(&sem->wait);
}
int
......@@ -86,10 +88,12 @@ __down_interruptible (struct semaphore * sem)
int retval = 0;
struct task_struct *tsk = current;
DECLARE_WAITQUEUE(wait, tsk);
unsigned long flags;
tsk->state = TASK_INTERRUPTIBLE;
add_wait_queue_exclusive(&sem->wait, &wait);
spin_lock_irqsave(&sem->wait.lock, flags);
add_wait_queue_exclusive_locked(&sem->wait, &wait);
spin_lock_irq(&semaphore_lock);
sem->sleepers ++;
for (;;) {
int sleepers = sem->sleepers;
......@@ -110,25 +114,27 @@ __down_interruptible (struct semaphore * sem)
/*
* Add "everybody else" into it. They aren't
* playing, because we own the spinlock. The
* "-1" is because we're still hoping to get
* the lock.
* playing, because we own the spinlock in
* wait_queue_head. The "-1" is because we're
* still hoping to get the semaphore.
*/
if (!atomic_add_negative(sleepers - 1, &sem->count)) {
sem->sleepers = 0;
break;
}
sem->sleepers = 1; /* us - see -1 above */
spin_unlock_irq(&semaphore_lock);
spin_unlock_irqrestore(&sem->wait.lock, flags);
schedule();
spin_lock_irqsave(&sem->wait.lock, flags);
tsk->state = TASK_INTERRUPTIBLE;
spin_lock_irq(&semaphore_lock);
}
spin_unlock_irq(&semaphore_lock);
remove_wait_queue_locked(&sem->wait, &wait);
wake_up_locked(&sem->wait);
spin_unlock_irqrestore(&sem->wait.lock, flags);
tsk->state = TASK_RUNNING;
remove_wait_queue(&sem->wait, &wait);
wake_up(&sem->wait);
return retval;
}
......@@ -142,17 +148,19 @@ __down_trylock (struct semaphore *sem)
unsigned long flags;
int sleepers;
spin_lock_irqsave(&semaphore_lock, flags);
spin_lock_irqsave(&sem->wait.lock, flags);
sleepers = sem->sleepers + 1;
sem->sleepers = 0;
/*
* Add "everybody else" and us into it. They aren't
* playing, because we own the spinlock.
* playing, because we own the spinlock in the
* wait_queue_head.
*/
if (!atomic_add_negative(sleepers, &sem->count))
wake_up(&sem->wait);
if (!atomic_add_negative(sleepers, &sem->count)) {
wake_up_locked(&sem->wait);
}
spin_unlock_irqrestore(&semaphore_lock, flags);
spin_unlock_irqrestore(&sem->wait.lock, flags);
return 1;
}
......@@ -354,6 +354,15 @@ setup_sigcontext (struct sigcontext *sc, sigset_t *mask, struct sigscratch *scr)
return err;
}
/*
* Check whether the register-backing store is already on the signal stack.
*/
static inline int
rbs_on_sig_stack (unsigned long bsp)
{
return (bsp - current->sas_ss_sp < current->sas_ss_size);
}
static long
setup_frame (int sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *set,
struct sigscratch *scr)
......@@ -366,10 +375,17 @@ setup_frame (int sig, struct k_sigaction *ka, siginfo_t *info, sigset_t *set,
frame = (void *) scr->pt.r12;
tramp_addr = GATE_ADDR + (ia64_sigtramp - __start_gate_section);
if ((ka->sa.sa_flags & SA_ONSTACK) != 0 && !on_sig_stack((unsigned long) frame)) {
new_rbs = (current->sas_ss_sp + sizeof(long) - 1) & ~(sizeof(long) - 1);
frame = (void *) ((current->sas_ss_sp + current->sas_ss_size)
& ~(STACK_ALIGN - 1));
if (ka->sa.sa_flags & SA_ONSTACK) {
/*
* We need to check the memory and register stacks separately, because
* they're switched separately (memory stack is switched in the kernel,
* register stack is switched in the signal trampoline).
*/
if (!on_sig_stack((unsigned long) frame))
frame = (void *) ((current->sas_ss_sp + current->sas_ss_size)
& ~(STACK_ALIGN - 1));
if (!rbs_on_sig_stack(scr->pt.ar_bspstore))
new_rbs = (current->sas_ss_sp + sizeof(long) - 1) & ~(sizeof(long) - 1);
}
frame = (void *) frame - ((sizeof(*frame) + STACK_ALIGN - 1) & ~(STACK_ALIGN - 1));
......@@ -460,7 +476,6 @@ handle_signal (unsigned long sig, struct k_sigaction *ka, siginfo_t *info, sigse
long
ia64_do_signal (sigset_t *oldset, struct sigscratch *scr, long in_syscall)
{
struct signal_struct *sig;
struct k_sigaction *ka;
siginfo_t info;
long restart = in_syscall;
......@@ -487,7 +502,7 @@ ia64_do_signal (sigset_t *oldset, struct sigscratch *scr, long in_syscall)
}
} else
#endif
if (scr->pt.r10 != -1) {
if (scr->pt.r10 != -1)
/*
* A system calls has to be restarted only if one of the error codes
* ERESTARTNOHAND, ERESTARTSYS, or ERESTARTNOINTR is returned. If r10
......@@ -495,101 +510,14 @@ ia64_do_signal (sigset_t *oldset, struct sigscratch *scr, long in_syscall)
* restart the syscall, so we can clear the "restart" flag here.
*/
restart = 0;
}
for (;;) {
unsigned long signr;
spin_lock_irq(&current->sigmask_lock);
signr = dequeue_signal(&current->blocked, &info);
spin_unlock_irq(&current->sigmask_lock);
while (1) {
int signr = get_signal_to_deliver(&info, &scr->pt);
if (!signr)
if (signr <= 0)
break;
if ((current->ptrace & PT_PTRACED) && signr != SIGKILL) {
/* Let the debugger run. */
current->exit_code = signr;
current->thread.siginfo = &info;
current->state = TASK_STOPPED;
notify_parent(current, SIGCHLD);
schedule();
signr = current->exit_code;
current->thread.siginfo = 0;
/* We're back. Did the debugger cancel the sig? */
if (!signr)
continue;
current->exit_code = 0;
/* The debugger continued. Ignore SIGSTOP. */
if (signr == SIGSTOP)
continue;
/* Update the siginfo structure. Is this good? */
if (signr != info.si_signo) {
info.si_signo = signr;
info.si_errno = 0;
info.si_code = SI_USER;
info.si_pid = current->parent->pid;
info.si_uid = current->parent->uid;
}
/* If the (new) signal is now blocked, requeue it. */
if (sigismember(&current->blocked, signr)) {
send_sig_info(signr, &info, current);
continue;
}
}
ka = &current->sig->action[signr - 1];
if (ka->sa.sa_handler == SIG_IGN) {
if (signr != SIGCHLD)
continue;
/* Check for SIGCHLD: it's special. */
while (sys_wait4(-1, NULL, WNOHANG, NULL) > 0)
/* nothing */;
continue;
}
if (ka->sa.sa_handler == SIG_DFL) {
int exit_code = signr;
/* Init gets no signals it doesn't want. */
if (current->pid == 1)
continue;
switch (signr) {
case SIGCONT: case SIGCHLD: case SIGWINCH: case SIGURG:
continue;
case SIGTSTP: case SIGTTIN: case SIGTTOU:
if (is_orphaned_pgrp(current->pgrp))
continue;
/* FALLTHRU */
case SIGSTOP:
current->state = TASK_STOPPED;
current->exit_code = signr;
sig = current->parent->sig;
if (sig && !(sig->action[SIGCHLD-1].sa.sa_flags & SA_NOCLDSTOP))
notify_parent(current, SIGCHLD);
schedule();
continue;
case SIGQUIT: case SIGILL: case SIGTRAP:
case SIGABRT: case SIGFPE: case SIGSEGV:
case SIGBUS: case SIGSYS: case SIGXCPU: case SIGXFSZ:
if (do_coredump(signr, &scr->pt))
exit_code |= 0x80;
/* FALLTHRU */
default:
sig_exit(signr, exit_code, &info);
/* NOTREACHED */
}
}
if (restart) {
switch (errno) {
......@@ -601,7 +529,7 @@ ia64_do_signal (sigset_t *oldset, struct sigscratch *scr, long in_syscall)
scr->pt.r8 = -EINTR;
else
#endif
scr->pt.r8 = EINTR;
scr->pt.r8 = EINTR;
/* note: scr->pt.r10 is already -1 */
break;
}
......@@ -612,13 +540,14 @@ ia64_do_signal (sigset_t *oldset, struct sigscratch *scr, long in_syscall)
scr->pt.cr_iip -= 2;
} else
#endif
ia64_decrement_ip(&scr->pt);
ia64_decrement_ip(&scr->pt);
}
}
/* Whee! Actually deliver the signal. If the
delivery failed, we need to continue to iterate in
this loop so we can deliver the SIGSEGV... */
/*
* Whee! Actually deliver the signal. If the delivery failed, we need to
* continue to iterate in this loop so we can deliver the SIGSEGV...
*/
if (handle_signal(signr, ka, &info, oldset, scr))
return 1;
}
......@@ -634,9 +563,8 @@ ia64_do_signal (sigset_t *oldset, struct sigscratch *scr, long in_syscall)
} else
#endif
/*
* Note: the syscall number is in r15 which is
* saved in pt_regs so all we need to do here
* is adjust ip so that the "break"
* Note: the syscall number is in r15 which is saved in pt_regs so
* all we need to do here is adjust ip so that the "break"
* instruction gets re-executed.
*/
ia64_decrement_ip(&scr->pt);
......
......@@ -257,31 +257,6 @@ smp_setup_percpu_timer (void)
local_cpu_data->prof_multiplier = 1;
}
/*
* Architecture specific routine called by the kernel just before init is
* fired off. This allows the BP to have everything in order [we hope].
* At the end of this all the APs will hit the system scheduling and off
* we go. Each AP will jump through the kernel
* init into idle(). At this point the scheduler will one day take over
* and give them jobs to do. smp_callin is a standard routine
* we use to track CPUs as they power up.
*/
static volatile atomic_t smp_commenced = ATOMIC_INIT(0);
static void __init
smp_commence (void)
{
/*
* Lets the callins below out of their loop.
*/
Dprintk("Setting commenced=1, go go go\n");
wmb();
atomic_set(&smp_commenced, 1);
}
static void __init
smp_callin (void)
{
......@@ -361,7 +336,7 @@ fork_by_hand (void)
* don't care about the eip and regs settings since we'll never reschedule the
* forked task.
*/
return do_fork(CLONE_VM|CLONE_IDLETASK, 0, 0, 0);
return do_fork(CLONE_VM|CLONE_IDLETASK, 0, 0, 0, NULL);
}
static int __init
......
......@@ -19,6 +19,12 @@
#include <asm/shmparam.h>
#include <asm/uaccess.h>
#ifdef CONFIG_HUGETLB_PAGE
# define SHMLBA_HPAGE HPAGE_SIZE
# define COLOR_HALIGN(addr) (((addr) + SHMLBA_HPAGE - 1) & ~(SHMLBA_HPAGE - 1))
# define TASK_HPAGE_BASE ((REGION_HPAGE << REGION_SHIFT) | HPAGE_SIZE)
#endif
unsigned long
arch_get_unmapped_area (struct file *filp, unsigned long addr, unsigned long len,
unsigned long pgoff, unsigned long flags)
......@@ -56,16 +62,14 @@ arch_get_unmapped_area (struct file *filp, unsigned long addr, unsigned long len
}
asmlinkage long
ia64_getpriority (int which, int who, long arg2, long arg3, long arg4, long arg5, long arg6,
long arg7, long stack)
ia64_getpriority (int which, int who)
{
struct pt_regs *regs = (struct pt_regs *) &stack;
extern long sys_getpriority (int, int);
long prio;
prio = sys_getpriority(which, who);
if (prio >= 0) {
regs->r8 = 0; /* ensure negative priority is not mistaken as error code */
force_successful_syscall_return();
prio = 20 - prio;
}
return prio;
......@@ -79,10 +83,8 @@ sys_getpagesize (void)
}
asmlinkage unsigned long
ia64_shmat (int shmid, void *shmaddr, int shmflg, long arg3, long arg4, long arg5, long arg6,
long arg7, long stack)
ia64_shmat (int shmid, void *shmaddr, int shmflg)
{
struct pt_regs *regs = (struct pt_regs *) &stack;
unsigned long raddr;
int retval;
......@@ -90,16 +92,14 @@ ia64_shmat (int shmid, void *shmaddr, int shmflg, long arg3, long arg4, long arg
if (retval < 0)
return retval;
regs->r8 = 0; /* ensure negative addresses are not mistaken as an error code */
force_successful_syscall_return();
return raddr;
}
asmlinkage unsigned long
ia64_brk (unsigned long brk, long arg1, long arg2, long arg3,
long arg4, long arg5, long arg6, long arg7, long stack)
ia64_brk (unsigned long brk)
{
extern int vm_enough_memory (long pages);
struct pt_regs *regs = (struct pt_regs *) &stack;
unsigned long rlim, retval, newbrk, oldbrk;
struct mm_struct *mm = current->mm;
......@@ -145,7 +145,7 @@ ia64_brk (unsigned long brk, long arg1, long arg2, long arg3,
out:
retval = mm->brk;
up_write(&mm->mmap_sem);
regs->r8 = 0; /* ensure large retval isn't mistaken as error code */
force_successful_syscall_return();
return retval;
}
......@@ -222,32 +222,98 @@ out: if (file)
* of) files that are larger than the address space of the CPU.
*/
asmlinkage unsigned long
sys_mmap2 (unsigned long addr, unsigned long len, int prot, int flags, int fd, long pgoff,
long arg6, long arg7, long stack)
sys_mmap2 (unsigned long addr, unsigned long len, int prot, int flags, int fd, long pgoff)
{
struct pt_regs *regs = (struct pt_regs *) &stack;
addr = do_mmap2(addr, len, prot, flags, fd, pgoff);
if (!IS_ERR((void *) addr))
regs->r8 = 0; /* ensure large addresses are not mistaken as failures... */
force_successful_syscall_return();
return addr;
}
asmlinkage unsigned long
sys_mmap (unsigned long addr, unsigned long len, int prot, int flags,
int fd, long off, long arg6, long arg7, long stack)
sys_mmap (unsigned long addr, unsigned long len, int prot, int flags, int fd, long off)
{
struct pt_regs *regs = (struct pt_regs *) &stack;
if ((off & ~PAGE_MASK) != 0)
return -EINVAL;
addr = do_mmap2(addr, len, prot, flags, fd, off >> PAGE_SHIFT);
if (!IS_ERR((void *) addr))
regs->r8 = 0; /* ensure large addresses are not mistaken as failures... */
force_successful_syscall_return();
return addr;
}
#ifdef CONFIG_HUGETLB_PAGE
asmlinkage unsigned long
sys_alloc_hugepages (int key, unsigned long addr, size_t len, int prot, int flag)
{
struct mm_struct *mm = current->mm;
long retval;
extern int alloc_hugetlb_pages (int, unsigned long, unsigned long, int, int);
if ((key < 0) || (len & (HPAGE_SIZE - 1)))
return -EINVAL;
if (addr && ((REGION_NUMBER(addr) != REGION_HPAGE) || (addr & (HPAGE_SIZE - 1))))
addr = TASK_HPAGE_BASE;
if (!addr)
addr = TASK_HPAGE_BASE;
down_write(&mm->mmap_sem);
{
retval = arch_get_unmapped_area(NULL, COLOR_HALIGN(addr), len, 0, 0);
if (retval != -ENOMEM)
retval = alloc_hugetlb_pages(key, retval, len, prot, flag);
}
up_write(&mm->mmap_sem);
if (IS_ERR((void *) retval))
return retval;
force_successful_syscall_return();
return retval;
}
asmlinkage int
sys_free_hugepages (unsigned long addr)
{
struct mm_struct *mm = current->mm;
struct vm_area_struct *vma;
extern int free_hugepages(struct vm_area_struct *);
int retval;
vma = find_vma(mm, addr);
if (!vma || !is_vm_hugetlb_page(vma) || (vma->vm_start != addr))
return -EINVAL;
down_write(&mm->mmap_sem);
{
spin_lock(&mm->page_table_lock);
{
retval = free_hugepages(vma);
}
spin_unlock(&mm->page_table_lock);
}
up_write(&mm->mmap_sem);
return retval;
}
#else /* !CONFIG_HUGETLB_PAGE */
asmlinkage unsigned long
sys_alloc_hugepages (int key, size_t addr, unsigned long len, int prot, int flag)
{
return -ENOSYS;
}
asmlinkage unsigned long
sys_free_hugepages (unsigned long addr)
{
return -ENOSYS;
}
#endif /* !CONFIG_HUGETLB_PAGE */
asmlinkage long
sys_vm86 (long arg0, long arg1, long arg2, long arg3)
{
......@@ -256,16 +322,14 @@ sys_vm86 (long arg0, long arg1, long arg2, long arg3)
}
asmlinkage unsigned long
ia64_create_module (const char *name_user, size_t size, long arg2, long arg3,
long arg4, long arg5, long arg6, long arg7, long stack)
ia64_create_module (const char *name_user, size_t size)
{
extern unsigned long sys_create_module (const char *, size_t);
struct pt_regs *regs = (struct pt_regs *) &stack;
unsigned long addr;
addr = sys_create_module (name_user, size);
if (!IS_ERR((void *) addr))
regs->r8 = 0; /* ensure large addresses are not mistaken as failures... */
force_successful_syscall_return();
return addr;
}
......
......@@ -104,7 +104,8 @@ do_settimeofday (struct timeval *tv)
tv->tv_sec--;
}
xtime = *tv;
xtime.tv_sec = tv->tv_sec;
xtime.tv_nsec = 1000 * tv->tv_usec;
time_adjust = 0; /* stop active adjtime() */
time_status |= STA_UNSYNC;
time_maxerror = NTP_PHASE_LIMIT;
......@@ -135,7 +136,7 @@ do_gettimeofday (struct timeval *tv)
} while (cmpxchg(&last_time_offset, old, usec) != old);
sec = xtime.tv_sec;
usec += xtime.tv_usec;
usec += xtime.tv_nsec / 1000;
}
read_unlock_irqrestore(&xtime_lock, flags);
......
......@@ -15,6 +15,7 @@ obj-y := __divsi3.o __udivsi3.o __modsi3.o __umodsi3.o \
obj-$(CONFIG_ITANIUM) += copy_page.o copy_user.o memcpy.o
obj-$(CONFIG_MCKINLEY) += copy_page_mck.o memcpy_mck.o
obj-$(CONFIG_PERFMON) += carta_random.o
IGNORE_FLAGS_OBJS = __divsi3.o __udivsi3.o __modsi3.o __umodsi3.o \
__divdi3.o __udivdi3.o __moddi3.o __umoddi3.o
......
/*
* Fast, simple, yet decent quality random number generator based on
* a paper by David G. Carta ("Two Fast Implementations of the
* `Minimal Standard' Random Number Generator," Communications of the
* ACM, January, 1990).
*
* Copyright (C) 2002 Hewlett-Packard Co
* David Mosberger-Tang <davidm@hpl.hp.com>
*/
#include <asm/asmmacro.h>
#define a r2
#define m r3
#define lo r8
#define hi r9
#define t0 r16
#define t1 r17
#define seed r32
GLOBAL_ENTRY(carta_random32)
movl a = (16807 << 16) | 16807
;;
pmpyshr2.u t0 = a, seed, 0
pmpyshr2.u t1 = a, seed, 16
;;
unpack2.l t0 = t1, t0
dep m = -1, r0, 0, 31
;;
zxt4 lo = t0
shr.u hi = t0, 32
;;
dep t0 = 0, hi, 15, 49 // t0 = (hi & 0x7fff)
;;
shl t0 = t0, 16 // t0 = (hi & 0x7fff) << 16
shr t1 = hi, 15 // t1 = (hi >> 15)
;;
add lo = lo, t0
;;
cmp.gtu p6, p0 = lo, m
;;
(p6) and lo = lo, m
;;
(p6) add lo = 1, lo
;;
add lo = lo, t1
;;
cmp.gtu p6, p0 = lo, m
;;
(p6) and lo = lo, m
;;
(p6) add lo = 1, lo
br.ret.sptk.many rp
END(carta_random32)
......@@ -10,5 +10,6 @@
O_TARGET := mm.o
obj-y := init.o fault.o tlb.o extable.o
obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o
include $(TOPDIR)/Rules.make
This diff is collapsed.
......@@ -78,7 +78,7 @@ ia64_init_addr_space (void)
vma->vm_mm = current->mm;
vma->vm_start = IA64_RBS_BOT;
vma->vm_end = vma->vm_start + PAGE_SIZE;
vma->vm_page_prot = PAGE_COPY;
vma->vm_page_prot = protection_map[VM_READ | VM_WRITE];
vma->vm_flags = VM_READ|VM_WRITE|VM_MAYREAD|VM_MAYWRITE|VM_GROWSUP;
vma->vm_ops = NULL;
vma->vm_pgoff = 0;
......@@ -339,6 +339,14 @@ ia64_mmu_init (void *my_cpu_data)
/*
* Set up the page tables.
*/
#ifdef CONFIG_HUGETLB_PAGE
long htlbpagemem;
int htlbpage_max;
extern long htlbzone_pages;
extern struct list_head htlbpage_freelist;
#endif
void
paging_init (void)
{
......@@ -439,4 +447,29 @@ mem_init (void)
#ifdef CONFIG_IA32_SUPPORT
ia32_gdt_init();
#endif
#ifdef CONFIG_HUGETLB_PAGE
{
long i;
int j;
struct page *page, *map;
if ((htlbzone_pages << (HPAGE_SHIFT - PAGE_SHIFT)) >= max_low_pfn)
htlbzone_pages = (max_low_pfn >> ((HPAGE_SHIFT - PAGE_SHIFT) + 1));
INIT_LIST_HEAD(&htlbpage_freelist);
for (i = 0; i < htlbzone_pages; i++) {
page = alloc_pages(__GFP_HIGHMEM, HUGETLB_PAGE_ORDER);
if (!page)
break;
map = page;
for (j = 0; j < (HPAGE_SIZE/PAGE_SIZE); j++) {
SetPageReserved(map);
map++;
}
list_add(&page->list, &htlbpage_freelist);
}
printk("Total Huge_TLB_Page memory pages allocated %ld \n", i);
htlbzone_pages = htlbpagemem = i;
htlbpage_max = (int)i;
}
#endif
}
......@@ -63,7 +63,7 @@ wrap_mmu_context (struct mm_struct *mm)
read_lock(&tasklist_lock);
repeat:
for_each_task(tsk) {
for_each_process(tsk) {
if (!tsk->mm)
continue;
tsk_context = tsk->mm->context;
......
......@@ -9,6 +9,7 @@ BEGIN {
print " */"
print ""
print "#define CLONE_IDLETASK_BIT 12"
print "#define CLONE_SETTLS_BIT 19"
}
# look for .tab:
......
......@@ -202,6 +202,7 @@ main (int argc, char **argv)
}
printf ("\n#define CLONE_IDLETASK_BIT %ld\n", ia64_fls (CLONE_IDLETASK));
printf ("\n#define CLONE_SETTLS_BIT %ld\n", ia64_fls (CLONE_SETTLS));
printf ("\n#endif /* _ASM_IA64_OFFSETS_H */\n");
return 0;
......
......@@ -83,6 +83,7 @@
#define hardirq_trylock() (!in_interrupt())
#define hardirq_endlock() do { } while (0)
#define in_atomic() (preempt_count() != 0)
#define irq_enter() (preempt_count() += HARDIRQ_OFFSET)
#if CONFIG_PREEMPT
......
......@@ -90,20 +90,6 @@ ide_init_default_hwifs (void)
#endif
}
#define ide_request_irq(irq,hand,flg,dev,id) request_irq((irq),(hand),(flg),(dev),(id))
#define ide_free_irq(irq,dev_id) free_irq((irq), (dev_id))
#define ide_check_region(from,extent) check_region((from), (extent))
#define ide_request_region(from,extent,name) request_region((from), (extent), (name))
#define ide_release_region(from,extent) release_region((from), (extent))
/*
* The following are not needed for the non-m68k ports
*/
#define ide_ack_intr(hwif) (1)
#define ide_fix_driveid(id) do {} while (0)
#define ide_release_lock(lock) do {} while (0)
#define ide_get_lock(lock, hdlr, data) do {} while (0)
#endif /* __KERNEL__ */
#endif /* __ASM_IA64_IDE_H */
......@@ -110,6 +110,10 @@ reload_context (struct mm_struct *mm)
rr2 = rr0 + 2*rid_incr;
rr3 = rr0 + 3*rid_incr;
rr4 = rr0 + 4*rid_incr;
#ifdef CONFIG_HUGETLB_PAGE
rr4 = (rr4 & (~(0xfcUL))) | (HPAGE_SHIFT << 2);
#endif
ia64_set_rr(0x0000000000000000, rr0);
ia64_set_rr(0x2000000000000000, rr1);
ia64_set_rr(0x4000000000000000, rr2);
......
......@@ -30,6 +30,32 @@
#define PAGE_MASK (~(PAGE_SIZE - 1))
#define PAGE_ALIGN(addr) (((addr) + PAGE_SIZE - 1) & PAGE_MASK)
#ifdef CONFIG_HUGETLB_PAGE
# if defined(CONFIG_HUGETLB_PAGE_SIZE_4GB)
# define HPAGE_SHIFT 32
# elif defined(CONFIG_HUGETLB_PAGE_SIZE_256MB)
# define HPAGE_SHIFT 28
# elif defined(CONFIG_HUGETLB_PAGE_SIZE_64MB)
# define HPAGE_SHIFT 26
# elif defined(CONFIG_HUGETLB_PAGE_SIZE_16MB)
# define HPAGE_SHIFT 24
# elif defined(CONFIG_HUGETLB_PAGE_SIZE_4MB)
# define HPAGE_SHIFT 22
# elif defined(CONFIG_HUGETLB_PAGE_SIZE_1MB)
# define HPAGE_SHIFT 20
# elif defined(CONFIG_HUGETLB_PAGE_SIZE_256KB)
# define HPAGE_SHIFT 18
# else
# error Unsupported IA-64 HugeTLB Page Size!
# endif
# define REGION_HPAGE (4UL) /* note: this is hardcoded in mmu_context.h:reload_context()!*/
# define REGION_SHIFT 61
# define HPAGE_SIZE (__IA64_UL_CONST(1) << HPAGE_SHIFT)
# define HPAGE_MASK (~(HPAGE_SIZE - 1))
#endif /* CONFIG_HUGETLB_PAGE */
#ifdef __ASSEMBLY__
# define __pa(x) ((x) - PAGE_OFFSET)
# define __va(x) ((x) + PAGE_OFFSET)
......@@ -87,6 +113,12 @@ typedef union ia64_va {
#define REGION_SIZE REGION_NUMBER(1)
#define REGION_KERNEL 7
#ifdef CONFIG_HUGETLB_PAGE
# define htlbpage_to_page(x) ((REGION_NUMBER(x) << 61) \
| (REGION_OFFSET(x) >> (HPAGE_SHIFT-PAGE_SHIFT)))
# define HUGETLB_PAGE_ORDER (HPAGE_SHIFT - PAGE_SHIFT)
#endif
#if (__GNUC__ > 3) || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)
# define ia64_abort() __builtin_trap()
#else
......
......@@ -78,6 +78,7 @@
#ifndef __ASSEMBLY__
#include <linux/types.h>
#include <asm/fpu.h>
/*
* Data types needed to pass information into PAL procedures and
......@@ -649,12 +650,43 @@ extern struct ia64_pal_retval ia64_pal_call_static (u64, u64, u64, u64, u64);
extern struct ia64_pal_retval ia64_pal_call_stacked (u64, u64, u64, u64);
extern struct ia64_pal_retval ia64_pal_call_phys_static (u64, u64, u64, u64);
extern struct ia64_pal_retval ia64_pal_call_phys_stacked (u64, u64, u64, u64);
#define PAL_CALL(iprv,a0,a1,a2,a3) iprv = ia64_pal_call_static(a0, a1, a2, a3, 0)
#define PAL_CALL_IC_OFF(iprv,a0,a1,a2,a3) iprv = ia64_pal_call_static(a0, a1, a2, a3, 1)
#define PAL_CALL_STK(iprv,a0,a1,a2,a3) iprv = ia64_pal_call_stacked(a0, a1, a2, a3)
#define PAL_CALL_PHYS(iprv,a0,a1,a2,a3) iprv = ia64_pal_call_phys_static(a0, a1, a2, a3)
#define PAL_CALL_PHYS_STK(iprv,a0,a1,a2,a3) iprv = ia64_pal_call_phys_stacked(a0, a1, a2, a3)
extern void ia64_save_scratch_fpregs (struct ia64_fpreg *);
extern void ia64_load_scratch_fpregs (struct ia64_fpreg *);
#define PAL_CALL(iprv,a0,a1,a2,a3) do { \
struct ia64_fpreg fr[6]; \
ia64_save_scratch_fpregs(fr); \
iprv = ia64_pal_call_static(a0, a1, a2, a3, 0); \
ia64_load_scratch_fpregs(fr); \
} while (0)
#define PAL_CALL_IC_OFF(iprv,a0,a1,a2,a3) do { \
struct ia64_fpreg fr[6]; \
ia64_save_scratch_fpregs(fr); \
iprv = ia64_pal_call_static(a0, a1, a2, a3, 1); \
ia64_load_scratch_fpregs(fr); \
} while (0)
#define PAL_CALL_STK(iprv,a0,a1,a2,a3) do { \
struct ia64_fpreg fr[6]; \
ia64_save_scratch_fpregs(fr); \
iprv = ia64_pal_call_stacked(a0, a1, a2, a3); \
ia64_load_scratch_fpregs(fr); \
} while (0)
#define PAL_CALL_PHYS(iprv,a0,a1,a2,a3) do { \
struct ia64_fpreg fr[6]; \
ia64_save_scratch_fpregs(fr); \
iprv = ia64_pal_call_phys_static(a0, a1, a2, a3); \
ia64_load_scratch_fpregs(fr); \
} while (0)
#define PAL_CALL_PHYS_STK(iprv,a0,a1,a2,a3) do { \
struct ia64_fpreg fr[6]; \
ia64_save_scratch_fpregs(fr); \
iprv = ia64_pal_call_phys_stacked(a0, a1, a2, a3); \
ia64_load_scratch_fpregs(fr); \
} while (0)
typedef int (*ia64_pal_handler) (u64, ...);
extern ia64_pal_handler ia64_pal;
......
......@@ -45,6 +45,7 @@
* PMC flags
*/
#define PFM_REGFL_OVFL_NOTIFY 0x1 /* send notification on overflow */
#define PFM_REGFL_RANDOM 0x2 /* randomize sampling periods */
/*
* PMD/PMC/IBR/DBR return flags (ignored on input)
......@@ -86,8 +87,10 @@ typedef struct {
unsigned long reg_short_reset;/* reset after counter overflow (small) */
unsigned long reg_reset_pmds[4]; /* which other counters to reset on overflow */
unsigned long reg_random_seed; /* seed value when randomization is used */
unsigned long reg_random_mask; /* bitmask used to limit random value */
unsigned long reserved[16]; /* for future use */
unsigned long reserved[14]; /* for future use */
} pfarg_reg_t;
typedef struct {
......@@ -132,28 +135,28 @@ typedef struct {
#define PFM_VERSION_MINOR(x) ((x) & 0xffff)
/*
* Entry header in the sampling buffer.
* The header is directly followed with the PMDS saved in increasing index
* order: PMD4, PMD5, .... How many PMDs are present is determined by the
* user program during context creation.
* Entry header in the sampling buffer. The header is directly followed
* with the PMDs saved in increasing index order: PMD4, PMD5, .... How
* many PMDs are present is determined by the user program during
* context creation.
*
* XXX: in this version of the entry, only up to 64 registers can be recorded
* This should be enough for quite some time. Always check sampling format
* before parsing entries!
* XXX: in this version of the entry, only up to 64 registers can be
* recorded. This should be enough for quite some time. Always check
* sampling format before parsing entries!
*
* Inn the case where multiple counters have overflowed at the same time, the
* rate field indicate the initial value of the first PMD, based on the index.
* For instance, if PMD2 and PMD5 have ovewrflowed for this entry, the rate field
* will show the initial value of PMD2.
* In the case where multiple counters overflow at the same time, the
* last_reset_value member indicates the initial value of the PMD with
* the smallest index. For instance, if PMD2 and PMD5 have overflowed,
* the last_reset_value member contains the initial value of PMD2.
*/
typedef struct {
int pid; /* identification of process */
int cpu; /* which cpu was used */
unsigned long rate; /* initial value of overflowed counter */
unsigned long stamp; /* timestamp */
unsigned long ip; /* where did the overflow interrupt happened */
unsigned long regs; /* bitmask of which registers overflowed */
unsigned long period; /* sampling period used by overflowed counter (smallest pmd index) */
int pid; /* identification of process */
int cpu; /* which cpu was used */
unsigned long last_reset_value; /* initial value of counter that overflowed */
unsigned long stamp; /* timestamp */
unsigned long ip; /* where did the overflow interrupt happened */
unsigned long regs; /* bitmask of which registers overflowed */
unsigned long period; /* unused */
} perfmon_smpl_entry_t;
extern int perfmonctl(pid_t pid, int cmd, void *arg, int narg);
......
......@@ -236,7 +236,15 @@ struct thread_struct {
__u64 ssd; /* IA32 stack selector descriptor */
__u64 old_k1; /* old value of ar.k1 */
__u64 old_iob; /* old IOBase value */
# define INIT_THREAD_IA32 0, 0, 0x17800000037fULL, 0, 0, 0, 0, 0, 0,
# define INIT_THREAD_IA32 .eflag = 0, \
.fsr = 0, \
.fcr = 0x17800000037fULL, \
.fir = 0, \
.fdr = 0, \
.csd = 0, \
.ssd = 0, \
.old_k1 = 0, \
.old_iob = 0,
#else
# define INIT_THREAD_IA32
#endif /* CONFIG_IA32_SUPPORT */
......@@ -248,7 +256,13 @@ struct thread_struct {
atomic_t pfm_notifiers_check; /* when >0, will cleanup ctx_notify_task in tasklist */
atomic_t pfm_owners_check; /* when >0, will cleanup ctx_owner in tasklist */
void *pfm_smpl_buf_list; /* list of sampling buffers to vfree */
# define INIT_THREAD_PM {0, }, {0, }, 0, NULL, {0}, {0}, NULL,
# define INIT_THREAD_PM .pmc = {0, }, \
.pmd = {0, }, \
.pfm_ovfl_block_reset = 0, \
.pfm_context = NULL, \
.pfm_notifiers_check = { 0 }, \
.pfm_owners_check = { 0 }, \
.pfm_smpl_buf_list = NULL,
#else
# define INIT_THREAD_PM
#endif
......@@ -258,16 +272,17 @@ struct thread_struct {
};
#define INIT_THREAD { \
flags: 0, \
ksp: 0, \
map_base: DEFAULT_MAP_BASE, \
task_size: DEFAULT_TASK_SIZE, \
siginfo: 0, \
.flags = 0, \
.ksp = 0, \
.map_base = DEFAULT_MAP_BASE, \
.task_size = DEFAULT_TASK_SIZE, \
.siginfo = 0, \
.last_fph_cpu = 0, \
INIT_THREAD_IA32 \
INIT_THREAD_PM \
dbr: {0, }, \
ibr: {0, }, \
fph: {{{{0}}}, } \
.dbr = {0, }, \
.ibr = {0, }, \
.fph = {{{{0}}}, } \
}
#define start_thread(regs,new_ip,new_sp) do { \
......
......@@ -37,9 +37,9 @@ ia64_rse_rnat_addr (unsigned long *slot_addr)
}
/*
* Calcuate the number of registers in the dirty partition starting at
* BSPSTORE with a size of DIRTY bytes. This isn't simply DIRTY
* divided by eight because the 64th slot is used to store ar.rnat.
* Calculate the number of registers in the dirty partition starting at BSPSTORE and
* ending at BSP. This isn't simply (BSP-BSPSTORE)/8 because every 64th slot stores
* ar.rnat.
*/
static __inline__ unsigned long
ia64_rse_num_regs (unsigned long *bspstore, unsigned long *bsp)
......
......@@ -38,9 +38,12 @@ extern spinlock_t sal_lock;
# define SAL_CALL(result,args...) do { \
unsigned long flags; \
struct ia64_fpreg fr[6]; \
ia64_save_scratch_fpregs(fr); \
spin_lock_irqsave(&sal_lock, flags); \
__SAL_CALL(result,args); \
spin_unlock_irqrestore(&sal_lock, flags); \
ia64_load_scratch_fpregs(fr); \
} while (0)
#define SAL_SET_VECTORS 0x01000000
......
......@@ -66,6 +66,7 @@ typedef struct siginfo {
long _band; /* POLL_IN, POLL_OUT, POLL_MSG (XPG requires a "long") */
int _fd;
} _sigpoll;
/* SIGPROF */
struct {
pid_t _pid; /* which child */
......
......@@ -166,7 +166,6 @@ struct k_sigaction {
# include <asm/sigcontext.h>
#define HAVE_ARCH_GET_SIGNAL_TO_DELIVER
#define HAVE_ARCH_SYS_PAUSE
#endif /* __KERNEL__ */
......
......@@ -108,7 +108,8 @@ typedef struct {
} rwlock_t;
#define RW_LOCK_UNLOCKED (rwlock_t) { 0, 0 }
#define rwlock_init(x) do { *(x) = RW_LOCK_UNLOCKED; } while(0)
#define rwlock_init(x) do { *(x) = RW_LOCK_UNLOCKED; } while(0)
#define rwlock_is_locked(x) (*(volatile int *) (x) != 0)
#define _raw_read_lock(rw) \
do { \
......
/* dummy (must be non-empty to prevent prejudicial removal...) */
......@@ -148,8 +148,8 @@ do { \
"cmp.ne p6,p7=%1,r0;;" \
"(p6) ssm psr.i;" \
"(p7) rsm psr.i;;" \
"srlz.d" \
: "=&r" (old_psr) : "r"((psr) & IA64_PSR_I) \
"(p6) srlz.d" \
: "=r" (old_psr) : "r"((psr) & IA64_PSR_I) \
: "p6", "p7", "memory"); \
if ((old_psr & IA64_PSR_I) && !(psr & IA64_PSR_I)) { \
__asm__ ("mov %0=ip" : "=r"(ip)); \
......@@ -174,6 +174,13 @@ do { \
#define local_irq_enable() __asm__ __volatile__ (";; ssm psr.i;; srlz.d" ::: "memory")
#define local_save_flags(flags) __asm__ __volatile__ ("mov %0=psr" : "=r" (flags) :: "memory")
#define irqs_disabled() \
({ \
unsigned long flags; \
local_save_flags(flags); \
(flags & IA64_PSR_I) == 0; \
})
/*
* Force an unresolved reference if someone tries to use
* ia64_fetch_and_add() with a bad value.
......@@ -367,14 +374,14 @@ struct task_struct;
extern void ia64_save_extra (struct task_struct *task);
extern void ia64_load_extra (struct task_struct *task);
#if defined(CONFIG_SMP) && defined(CONFIG_PERFMON)
#ifdef CONFIG_PERFMON
DECLARE_PER_CPU(int, pfm_syst_wide);
# define PERFMON_IS_SYSWIDE() (get_cpu_var(pfm_syst_wide) != 0)
#else
# define PERFMON_IS_SYSWIDE() (0)
#endif
#define __switch_to(prev,next,last) do { \
#define __switch_to(prev,next,last) do { \
if (((prev)->thread.flags & (IA64_THREAD_DBG_VALID|IA64_THREAD_PM_VALID)) \
|| IS_IA32_PROCESS(ia64_task_regs(prev)) || PERFMON_IS_SYSWIDE()) \
ia64_save_extra(prev); \
......
......@@ -3,15 +3,19 @@
/*
* Copyright (C) 1998-2001 Hewlett-Packard Co
* Copyright (C) 1998-2001 David Mosberger-Tang <davidm@hpl.hp.com>
* David Mosberger-Tang <davidm@hpl.hp.com>
*/
/*
* 2001/01/18 davidm Removed CLOCK_TICK_RATE. It makes no sense on IA-64.
* Also removed cacheflush_time as it's entirely unused.
*/
#include <asm/processor.h>
typedef unsigned long cycles_t;
#define CLOCK_TICK_RATE 100000000
static inline cycles_t
get_cycles (void)
{
......
......@@ -199,4 +199,8 @@ tlb_remove_page (mmu_gather_t *tlb, struct page *page)
#define tlb_start_vma(tlb, vma) do { } while (0)
#define tlb_end_vma(tlb, vma) do { } while (0)
#define tlb_remove_tlb_entry(tlb, ptep, addr) __tlb_remove_tlb_entry(tlb, ptep, addr)
#define pte_free_tlb(tlb, ptep) __pte_free_tlb(tlb, ptep)
#define pmd_free_tlb(tlb, ptep) __pmd_free_tlb(tlb, ptep)
#endif /* _ASM_IA64_TLB_H */
......@@ -225,7 +225,7 @@
#define __NR_security 1233
#define __NR_alloc_hugepages 1234
#define __NR_free_hugepages 1235
/* 1236 currently unused */
#define __NR_exit_group 1236
/* 1237 currently unused */
#define __NR_io_setup 1238
#define __NR_io_destroy 1239
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment