linux/arch/powerpc/platforms/pseries/setup.c

734 lines
18 KiB
C
Raw Normal View History

/*
* 64-bit pSeries and RS/6000 setup code.
*
* Copyright (C) 1995 Linus Torvalds
* Adapted from 'alpha' version by Gary Thomas
* Modified by Cort Dougan (cort@cs.nmt.edu)
* Modified by PPC64 Team, IBM Corp
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version
* 2 of the License, or (at your option) any later version.
*/
/*
* bootup setup stuff..
*/
#include <linux/cpu.h>
#include <linux/errno.h>
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/stddef.h>
#include <linux/unistd.h>
#include <linux/user.h>
#include <linux/tty.h>
#include <linux/major.h>
#include <linux/interrupt.h>
#include <linux/reboot.h>
#include <linux/init.h>
#include <linux/ioport.h>
#include <linux/console.h>
#include <linux/pci.h>
#include <linux/utsname.h>
#include <linux/adb.h>
#include <linux/export.h>
#include <linux/delay.h>
#include <linux/irq.h>
#include <linux/seq_file.h>
#include <linux/root_dev.h>
#include <linux/cpuidle.h>
#include <linux/of.h>
#include <linux/kexec.h>
#include <asm/mmu.h>
#include <asm/processor.h>
#include <asm/io.h>
#include <asm/pgtable.h>
#include <asm/prom.h>
#include <asm/rtas.h>
#include <asm/pci-bridge.h>
#include <asm/iommu.h>
#include <asm/dma.h>
#include <asm/machdep.h>
#include <asm/irq.h>
#include <asm/time.h>
#include <asm/nvram.h>
#include <asm/pmc.h>
#include <asm/mpic.h>
#include <asm/xics.h>
#include <asm/ppc-pci.h>
#include <asm/i8259.h>
#include <asm/udbg.h>
#include <asm/smp.h>
#include <asm/firmware.h>
#include <asm/eeh.h>
#include "plpar_wrappers.h"
#include "pseries.h"
int CMO_PrPSP = -1;
int CMO_SecPSP = -1;
unsigned long CMO_PageSize = (ASM_CONST(1) << IOMMU_PAGE_SHIFT);
EXPORT_SYMBOL(CMO_PageSize);
int fwnmi_active; /* TRUE if an FWNMI handler is present */
2006-07-03 13:36:01 +02:00
static struct device_node *pSeries_mpic_node;
static void pSeries_show_cpuinfo(struct seq_file *m)
{
struct device_node *root;
const char *model = "";
root = of_find_node_by_path("/");
if (root)
model = of_get_property(root, "model", NULL);
seq_printf(m, "machine\t\t: CHRP %s\n", model);
of_node_put(root);
}
/* Initialize firmware assisted non-maskable interrupts if
* the firmware supports this feature.
*/
static void __init fwnmi_init(void)
{
unsigned long system_reset_addr, machine_check_addr;
int ibm_nmi_register = rtas_token("ibm,nmi-register");
if (ibm_nmi_register == RTAS_UNKNOWN_SERVICE)
return;
/* If the kernel's not linked at zero we point the firmware at low
* addresses anyway, and use a trampoline to get to the real code. */
system_reset_addr = __pa(system_reset_fwnmi) - PHYSICAL_START;
machine_check_addr = __pa(machine_check_fwnmi) - PHYSICAL_START;
if (0 == rtas_call(ibm_nmi_register, 2, 1, NULL, system_reset_addr,
machine_check_addr))
fwnmi_active = 1;
}
static void pseries_8259_cascade(unsigned int irq, struct irq_desc *desc)
{
struct irq_chip *chip = irq_desc_get_chip(desc);
unsigned int cascade_irq = i8259_irq();
2006-07-03 13:36:01 +02:00
if (cascade_irq != NO_IRQ)
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 15:55:46 +02:00
generic_handle_irq(cascade_irq);
chip->irq_eoi(&desc->irq_data);
}
static void __init pseries_setup_i8259_cascade(void)
{
struct device_node *np, *old, *found = NULL;
unsigned int cascade;
const u32 *addrp;
unsigned long intack = 0;
int naddr;
for_each_node_by_type(np, "interrupt-controller") {
if (of_device_is_compatible(np, "chrp,iic")) {
found = np;
break;
}
}
if (found == NULL) {
printk(KERN_DEBUG "pic: no ISA interrupt controller\n");
return;
}
cascade = irq_of_parse_and_map(found, 0);
if (cascade == NO_IRQ) {
printk(KERN_ERR "pic: failed to map cascade interrupt");
return;
}
pr_debug("pic: cascade mapped to irq %d\n", cascade);
for (old = of_node_get(found); old != NULL ; old = np) {
np = of_get_parent(old);
of_node_put(old);
if (np == NULL)
break;
if (strcmp(np->name, "pci") != 0)
continue;
addrp = of_get_property(np, "8259-interrupt-acknowledge", NULL);
if (addrp == NULL)
continue;
naddr = of_n_addr_cells(np);
intack = addrp[naddr-1];
if (naddr > 1)
intack |= ((unsigned long)addrp[naddr-2]) << 32;
}
if (intack)
printk(KERN_DEBUG "pic: PCI 8259 intack at 0x%016lx\n", intack);
i8259_init(found, intack);
of_node_put(found);
irq_set_chained_handler(cascade, pseries_8259_cascade);
}
2006-07-03 13:36:01 +02:00
static void __init pseries_mpic_init_IRQ(void)
{
struct device_node *np;
const unsigned int *opprop;
unsigned long openpic_addr = 0;
2006-07-03 13:36:01 +02:00
int naddr, n, i, opplen;
struct mpic *mpic;
2006-07-03 13:36:01 +02:00
np = of_find_node_by_path("/");
naddr = of_n_addr_cells(np);
opprop = of_get_property(np, "platform-open-pic", &opplen);
if (opprop != 0) {
2006-07-03 13:36:01 +02:00
openpic_addr = of_read_number(opprop, naddr);
printk(KERN_DEBUG "OpenPIC addr: %lx\n", openpic_addr);
}
2006-07-03 13:36:01 +02:00
of_node_put(np);
BUG_ON(openpic_addr == 0);
/* Setup the openpic driver */
mpic = mpic_alloc(pSeries_mpic_node, openpic_addr,
MPIC_NO_RESET, 16, 0, " MPIC ");
2006-07-03 13:36:01 +02:00
BUG_ON(mpic == NULL);
/* Add ISUs */
opplen /= sizeof(u32);
for (n = 0, i = naddr; i < opplen; i += naddr, n++) {
unsigned long isuaddr = of_read_number(opprop + i, naddr);
mpic_assign_isu(mpic, n, isuaddr);
}
/* Setup top-level get_irq */
ppc_md.get_irq = mpic_get_irq;
2006-07-03 13:36:01 +02:00
/* All ISUs are setup, complete initialization */
mpic_init(mpic);
/* Look for cascade */
pseries_setup_i8259_cascade();
}
static void __init pseries_xics_init_IRQ(void)
{
xics_init();
pseries_setup_i8259_cascade();
}
static void pseries_lpar_enable_pmcs(void)
{
unsigned long set, reset;
set = 1UL << 63;
reset = 0;
plpar_hcall_norets(H_PERFMON, set, reset);
}
2006-07-03 13:36:01 +02:00
static void __init pseries_discover_pic(void)
{
struct device_node *np;
const char *typep;
2006-07-03 13:36:01 +02:00
for (np = NULL; (np = of_find_node_by_name(np,
"interrupt-controller"));) {
typep = of_get_property(np, "compatible", NULL);
2006-07-03 13:36:01 +02:00
if (strstr(typep, "open-pic")) {
pSeries_mpic_node = of_node_get(np);
ppc_md.init_IRQ = pseries_mpic_init_IRQ;
setup_kexec_cpu_down_mpic();
2006-07-03 13:36:01 +02:00
smp_init_pseries_mpic();
return;
} else if (strstr(typep, "ppc-xicp")) {
ppc_md.init_IRQ = pseries_xics_init_IRQ;
setup_kexec_cpu_down_xics();
2006-07-03 13:36:01 +02:00
smp_init_pseries_xics();
return;
}
}
printk(KERN_ERR "pSeries_discover_pic: failed to recognize"
" interrupt-controller\n");
}
static int pci_dn_reconfig_notifier(struct notifier_block *nb, unsigned long action, void *node)
{
struct device_node *np = node;
struct pci_dn *pci = NULL;
int err = NOTIFY_OK;
switch (action) {
case OF_RECONFIG_ATTACH_NODE:
pci = np->parent->data;
if (pci) {
update_dn_pci_info(np, pci->phb);
/* Create EEH device for the OF node */
eeh_dev_init(np, pci->phb);
}
break;
default:
err = NOTIFY_DONE;
break;
}
return err;
}
static struct notifier_block pci_dn_reconfig_nb = {
.notifier_call = pci_dn_reconfig_notifier,
};
struct kmem_cache *dtl_cache;
cputime: Generic on-demand virtual cputime accounting If we want to stop the tick further idle, we need to be able to account the cputime without using the tick. Virtual based cputime accounting solves that problem by hooking into kernel/user boundaries. However implementing CONFIG_VIRT_CPU_ACCOUNTING require low level hooks and involves more overhead. But we already have a generic context tracking subsystem that is required for RCU needs by archs which plan to shut down the tick outside idle. This patch implements a generic virtual based cputime accounting that relies on these generic kernel/user hooks. There are some upsides of doing this: - This requires no arch code to implement CONFIG_VIRT_CPU_ACCOUNTING if context tracking is already built (already necessary for RCU in full tickless mode). - We can rely on the generic context tracking subsystem to dynamically (de)activate the hooks, so that we can switch anytime between virtual and tick based accounting. This way we don't have the overhead of the virtual accounting when the tick is running periodically. And one downside: - There is probably more overhead than a native virtual based cputime accounting. But this relies on hooks that are already set anyway. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Li Zhong <zhong@linux.vnet.ibm.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de>
2012-07-25 07:56:04 +02:00
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
powerpc: Account time using timebase rather than PURR Currently, when CONFIG_VIRT_CPU_ACCOUNTING is enabled, we use the PURR register for measuring the user and system time used by processes, as well as other related times such as hardirq and softirq times. This turns out to be quite confusing for users because it means that a program will often be measured as taking less time when run on a multi-threaded processor (SMT2 or SMT4 mode) than it does when run on a single-threaded processor (ST mode), even though the program takes longer to finish. The discrepancy is accounted for as stolen time, which is also confusing, particularly when there are no other partitions running. This changes the accounting to use the timebase instead, meaning that the reported user and system times are the actual number of real-time seconds that the program was executing on the processor thread, regardless of which SMT mode the processor is in. Thus a program will generally show greater user and system times when run on a multi-threaded processor than on a single-threaded processor. On pSeries systems on POWER5 or later processors, we measure the stolen time (time when this partition wasn't running) using the hypervisor dispatch trace log. We check for new entries in the log on every entry from user mode and on every transition from kernel process context to soft or hard IRQ context (i.e. when account_system_vtime() gets called). So that we can correctly distinguish time stolen from user time and time stolen from system time, without having to check the log on every exit to user mode, we store separate timestamps for exit to user mode and entry from user mode. On systems that have a SPURR (POWER6 and POWER7), we read the SPURR in account_system_vtime() (as before), and then apportion the SPURR ticks since the last time we read it between scaled user time and scaled system time according to the relative proportions of user time and system time over the same interval. This avoids having to read the SPURR on every kernel entry and exit. On systems that have PURR but not SPURR (i.e., POWER5), we do the same using the PURR rather than the SPURR. This disables the DTL user interface in /sys/debug/kernel/powerpc/dtl for now since it conflicts with the use of the dispatch trace log by the time accounting code. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2010-08-26 21:56:43 +02:00
/*
* Allocate space for the dispatch trace log for all possible cpus
* and register the buffers with the hypervisor. This is used for
* computing time stolen by the hypervisor.
*/
static int alloc_dispatch_logs(void)
{
int cpu, ret;
struct paca_struct *pp;
struct dtl_entry *dtl;
if (!firmware_has_feature(FW_FEATURE_SPLPAR))
return 0;
if (!dtl_cache)
return 0;
powerpc: Account time using timebase rather than PURR Currently, when CONFIG_VIRT_CPU_ACCOUNTING is enabled, we use the PURR register for measuring the user and system time used by processes, as well as other related times such as hardirq and softirq times. This turns out to be quite confusing for users because it means that a program will often be measured as taking less time when run on a multi-threaded processor (SMT2 or SMT4 mode) than it does when run on a single-threaded processor (ST mode), even though the program takes longer to finish. The discrepancy is accounted for as stolen time, which is also confusing, particularly when there are no other partitions running. This changes the accounting to use the timebase instead, meaning that the reported user and system times are the actual number of real-time seconds that the program was executing on the processor thread, regardless of which SMT mode the processor is in. Thus a program will generally show greater user and system times when run on a multi-threaded processor than on a single-threaded processor. On pSeries systems on POWER5 or later processors, we measure the stolen time (time when this partition wasn't running) using the hypervisor dispatch trace log. We check for new entries in the log on every entry from user mode and on every transition from kernel process context to soft or hard IRQ context (i.e. when account_system_vtime() gets called). So that we can correctly distinguish time stolen from user time and time stolen from system time, without having to check the log on every exit to user mode, we store separate timestamps for exit to user mode and entry from user mode. On systems that have a SPURR (POWER6 and POWER7), we read the SPURR in account_system_vtime() (as before), and then apportion the SPURR ticks since the last time we read it between scaled user time and scaled system time according to the relative proportions of user time and system time over the same interval. This avoids having to read the SPURR on every kernel entry and exit. On systems that have PURR but not SPURR (i.e., POWER5), we do the same using the PURR rather than the SPURR. This disables the DTL user interface in /sys/debug/kernel/powerpc/dtl for now since it conflicts with the use of the dispatch trace log by the time accounting code. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2010-08-26 21:56:43 +02:00
for_each_possible_cpu(cpu) {
pp = &paca[cpu];
dtl = kmem_cache_alloc(dtl_cache, GFP_KERNEL);
powerpc: Account time using timebase rather than PURR Currently, when CONFIG_VIRT_CPU_ACCOUNTING is enabled, we use the PURR register for measuring the user and system time used by processes, as well as other related times such as hardirq and softirq times. This turns out to be quite confusing for users because it means that a program will often be measured as taking less time when run on a multi-threaded processor (SMT2 or SMT4 mode) than it does when run on a single-threaded processor (ST mode), even though the program takes longer to finish. The discrepancy is accounted for as stolen time, which is also confusing, particularly when there are no other partitions running. This changes the accounting to use the timebase instead, meaning that the reported user and system times are the actual number of real-time seconds that the program was executing on the processor thread, regardless of which SMT mode the processor is in. Thus a program will generally show greater user and system times when run on a multi-threaded processor than on a single-threaded processor. On pSeries systems on POWER5 or later processors, we measure the stolen time (time when this partition wasn't running) using the hypervisor dispatch trace log. We check for new entries in the log on every entry from user mode and on every transition from kernel process context to soft or hard IRQ context (i.e. when account_system_vtime() gets called). So that we can correctly distinguish time stolen from user time and time stolen from system time, without having to check the log on every exit to user mode, we store separate timestamps for exit to user mode and entry from user mode. On systems that have a SPURR (POWER6 and POWER7), we read the SPURR in account_system_vtime() (as before), and then apportion the SPURR ticks since the last time we read it between scaled user time and scaled system time according to the relative proportions of user time and system time over the same interval. This avoids having to read the SPURR on every kernel entry and exit. On systems that have PURR but not SPURR (i.e., POWER5), we do the same using the PURR rather than the SPURR. This disables the DTL user interface in /sys/debug/kernel/powerpc/dtl for now since it conflicts with the use of the dispatch trace log by the time accounting code. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2010-08-26 21:56:43 +02:00
if (!dtl) {
pr_warn("Failed to allocate dispatch trace log for cpu %d\n",
cpu);
pr_warn("Stolen time statistics will be unreliable\n");
break;
}
pp->dtl_ridx = 0;
pp->dispatch_log = dtl;
pp->dispatch_log_end = dtl + N_DISPATCH_LOG;
pp->dtl_curr = dtl;
}
/* Register the DTL for the current (boot) cpu */
dtl = get_paca()->dispatch_log;
get_paca()->dtl_ridx = 0;
get_paca()->dtl_curr = dtl;
get_paca()->lppaca_ptr->dtl_idx = 0;
/* hypervisor reads buffer length from this field */
dtl->enqueue_to_dispatch_time = DISPATCH_LOG_BYTES;
ret = register_dtl(hard_smp_processor_id(), __pa(dtl));
if (ret)
pr_err("WARNING: DTL registration of cpu %d (hw %d) failed "
"with %d\n", smp_processor_id(),
hard_smp_processor_id(), ret);
powerpc: Account time using timebase rather than PURR Currently, when CONFIG_VIRT_CPU_ACCOUNTING is enabled, we use the PURR register for measuring the user and system time used by processes, as well as other related times such as hardirq and softirq times. This turns out to be quite confusing for users because it means that a program will often be measured as taking less time when run on a multi-threaded processor (SMT2 or SMT4 mode) than it does when run on a single-threaded processor (ST mode), even though the program takes longer to finish. The discrepancy is accounted for as stolen time, which is also confusing, particularly when there are no other partitions running. This changes the accounting to use the timebase instead, meaning that the reported user and system times are the actual number of real-time seconds that the program was executing on the processor thread, regardless of which SMT mode the processor is in. Thus a program will generally show greater user and system times when run on a multi-threaded processor than on a single-threaded processor. On pSeries systems on POWER5 or later processors, we measure the stolen time (time when this partition wasn't running) using the hypervisor dispatch trace log. We check for new entries in the log on every entry from user mode and on every transition from kernel process context to soft or hard IRQ context (i.e. when account_system_vtime() gets called). So that we can correctly distinguish time stolen from user time and time stolen from system time, without having to check the log on every exit to user mode, we store separate timestamps for exit to user mode and entry from user mode. On systems that have a SPURR (POWER6 and POWER7), we read the SPURR in account_system_vtime() (as before), and then apportion the SPURR ticks since the last time we read it between scaled user time and scaled system time according to the relative proportions of user time and system time over the same interval. This avoids having to read the SPURR on every kernel entry and exit. On systems that have PURR but not SPURR (i.e., POWER5), we do the same using the PURR rather than the SPURR. This disables the DTL user interface in /sys/debug/kernel/powerpc/dtl for now since it conflicts with the use of the dispatch trace log by the time accounting code. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2010-08-26 21:56:43 +02:00
get_paca()->lppaca_ptr->dtl_enable_mask = 2;
return 0;
}
cputime: Generic on-demand virtual cputime accounting If we want to stop the tick further idle, we need to be able to account the cputime without using the tick. Virtual based cputime accounting solves that problem by hooking into kernel/user boundaries. However implementing CONFIG_VIRT_CPU_ACCOUNTING require low level hooks and involves more overhead. But we already have a generic context tracking subsystem that is required for RCU needs by archs which plan to shut down the tick outside idle. This patch implements a generic virtual based cputime accounting that relies on these generic kernel/user hooks. There are some upsides of doing this: - This requires no arch code to implement CONFIG_VIRT_CPU_ACCOUNTING if context tracking is already built (already necessary for RCU in full tickless mode). - We can rely on the generic context tracking subsystem to dynamically (de)activate the hooks, so that we can switch anytime between virtual and tick based accounting. This way we don't have the overhead of the virtual accounting when the tick is running periodically. And one downside: - There is probably more overhead than a native virtual based cputime accounting. But this relies on hooks that are already set anyway. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Li Zhong <zhong@linux.vnet.ibm.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de>
2012-07-25 07:56:04 +02:00
#else /* !CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
static inline int alloc_dispatch_logs(void)
{
return 0;
}
cputime: Generic on-demand virtual cputime accounting If we want to stop the tick further idle, we need to be able to account the cputime without using the tick. Virtual based cputime accounting solves that problem by hooking into kernel/user boundaries. However implementing CONFIG_VIRT_CPU_ACCOUNTING require low level hooks and involves more overhead. But we already have a generic context tracking subsystem that is required for RCU needs by archs which plan to shut down the tick outside idle. This patch implements a generic virtual based cputime accounting that relies on these generic kernel/user hooks. There are some upsides of doing this: - This requires no arch code to implement CONFIG_VIRT_CPU_ACCOUNTING if context tracking is already built (already necessary for RCU in full tickless mode). - We can rely on the generic context tracking subsystem to dynamically (de)activate the hooks, so that we can switch anytime between virtual and tick based accounting. This way we don't have the overhead of the virtual accounting when the tick is running periodically. And one downside: - There is probably more overhead than a native virtual based cputime accounting. But this relies on hooks that are already set anyway. Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Li Zhong <zhong@linux.vnet.ibm.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de>
2012-07-25 07:56:04 +02:00
#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
powerpc: Account time using timebase rather than PURR Currently, when CONFIG_VIRT_CPU_ACCOUNTING is enabled, we use the PURR register for measuring the user and system time used by processes, as well as other related times such as hardirq and softirq times. This turns out to be quite confusing for users because it means that a program will often be measured as taking less time when run on a multi-threaded processor (SMT2 or SMT4 mode) than it does when run on a single-threaded processor (ST mode), even though the program takes longer to finish. The discrepancy is accounted for as stolen time, which is also confusing, particularly when there are no other partitions running. This changes the accounting to use the timebase instead, meaning that the reported user and system times are the actual number of real-time seconds that the program was executing on the processor thread, regardless of which SMT mode the processor is in. Thus a program will generally show greater user and system times when run on a multi-threaded processor than on a single-threaded processor. On pSeries systems on POWER5 or later processors, we measure the stolen time (time when this partition wasn't running) using the hypervisor dispatch trace log. We check for new entries in the log on every entry from user mode and on every transition from kernel process context to soft or hard IRQ context (i.e. when account_system_vtime() gets called). So that we can correctly distinguish time stolen from user time and time stolen from system time, without having to check the log on every exit to user mode, we store separate timestamps for exit to user mode and entry from user mode. On systems that have a SPURR (POWER6 and POWER7), we read the SPURR in account_system_vtime() (as before), and then apportion the SPURR ticks since the last time we read it between scaled user time and scaled system time according to the relative proportions of user time and system time over the same interval. This avoids having to read the SPURR on every kernel entry and exit. On systems that have PURR but not SPURR (i.e., POWER5), we do the same using the PURR rather than the SPURR. This disables the DTL user interface in /sys/debug/kernel/powerpc/dtl for now since it conflicts with the use of the dispatch trace log by the time accounting code. Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2010-08-26 21:56:43 +02:00
static int alloc_dispatch_log_kmem_cache(void)
{
dtl_cache = kmem_cache_create("dtl", DISPATCH_LOG_BYTES,
DISPATCH_LOG_BYTES, 0, NULL);
if (!dtl_cache) {
pr_warn("Failed to create dispatch trace log buffer cache\n");
pr_warn("Stolen time statistics will be unreliable\n");
return 0;
}
return alloc_dispatch_logs();
}
early_initcall(alloc_dispatch_log_kmem_cache);
static void pSeries_idle(void)
{
/* This would call on the cpuidle framework, and the back-end pseries
* driver to go to idle states
*/
if (cpuidle_idle_call()) {
/* On error, execute default handler
* to go into low thread priority and possibly
* low power mode.
*/
HMT_low();
HMT_very_low();
}
}
/*
* Enable relocation on during exceptions. This has partition wide scope and
* may take a while to complete, if it takes longer than one second we will
* just give up rather than wasting any more time on this - if that turns out
* to ever be a problem in practice we can move this into a kernel thread to
* finish off the process later in boot.
*/
static int __init pSeries_enable_reloc_on_exc(void)
{
long rc;
unsigned int delay, total_delay = 0;
while (1) {
rc = enable_reloc_on_exceptions();
if (!H_IS_LONG_BUSY(rc))
return rc;
delay = get_longbusy_msecs(rc);
total_delay += delay;
if (total_delay > 1000) {
pr_warn("Warning: Giving up waiting to enable "
"relocation on exceptions (%u msec)!\n",
total_delay);
return rc;
}
mdelay(delay);
}
}
#ifdef CONFIG_KEXEC
static long pSeries_disable_reloc_on_exc(void)
{
long rc;
while (1) {
rc = disable_reloc_on_exceptions();
if (!H_IS_LONG_BUSY(rc))
return rc;
mdelay(get_longbusy_msecs(rc));
}
}
static void pSeries_machine_kexec(struct kimage *image)
{
long rc;
if (firmware_has_feature(FW_FEATURE_SET_MODE) &&
(image->type != KEXEC_TYPE_CRASH)) {
rc = pSeries_disable_reloc_on_exc();
if (rc != H_SUCCESS)
pr_warning("Warning: Failed to disable relocation on "
"exceptions: %ld\n", rc);
}
default_machine_kexec(image);
}
#endif
2006-07-03 13:36:01 +02:00
static void __init pSeries_setup_arch(void)
{
panic_timeout = 10;
2006-07-03 13:36:01 +02:00
/* Discover PIC type and setup ppc_md accordingly */
pseries_discover_pic();
/* openpic global configuration register (64-bit format). */
/* openpic Interrupt Source Unit pointer (64-bit format). */
/* python0 facility area (mmio) (64-bit format) REAL address. */
/* init to some ~sane value until calibrate_delay() runs */
loops_per_jiffy = 50000000;
fwnmi_init();
/* By default, only probe PCI (can be overriden by rtas_pci) */
pci_add_flags(PCI_PROBE_ONLY);
/* Find and initialize PCI host bridges */
init_pci_config_tokens();
find_and_init_phbs();
of_reconfig_notifier_register(&pci_dn_reconfig_nb);
pSeries_nvram_init();
if (firmware_has_feature(FW_FEATURE_SPLPAR)) {
vpa_init(boot_cpuid);
ppc_md.power_save = pSeries_idle;
}
if (firmware_has_feature(FW_FEATURE_LPAR))
ppc_md.enable_pmcs = pseries_lpar_enable_pmcs;
else
ppc_md.enable_pmcs = power4_enable_pmcs;
if (firmware_has_feature(FW_FEATURE_SET_MODE)) {
long rc;
if ((rc = pSeries_enable_reloc_on_exc()) != H_SUCCESS) {
pr_warn("Unable to enable relocation on exceptions: "
"%ld\n", rc);
}
}
}
static int __init pSeries_init_panel(void)
{
/* Manually leave the kernel version on the panel. */
ppc_md.progress("Linux ppc64\n", 0);
ppc_md.progress(init_utsname()->version, 0);
return 0;
}
machine_arch_initcall(pseries, pSeries_init_panel);
static int pseries_set_dabr(unsigned long dabr, unsigned long dabrx)
{
return plpar_hcall_norets(H_SET_DABR, dabr);
}
static int pseries_set_xdabr(unsigned long dabr, unsigned long dabrx)
{
/* Have to set at least one bit in the DABRX according to PAPR */
if (dabrx == 0 && dabr == 0)
dabrx = DABRX_USER;
/* PAPR says we can only set kernel and user bits */
dabrx &= DABRX_KERNEL | DABRX_USER;
return plpar_hcall_norets(H_SET_XDABR, dabr, dabrx);
}
#define CMO_CHARACTERISTICS_TOKEN 44
#define CMO_MAXLENGTH 1026
void pSeries_coalesce_init(void)
{
struct hvcall_mpp_x_data mpp_x_data;
if (firmware_has_feature(FW_FEATURE_CMO) && !h_get_mpp_x(&mpp_x_data))
powerpc_firmware_features |= FW_FEATURE_XCMO;
else
powerpc_firmware_features &= ~FW_FEATURE_XCMO;
}
/**
* fw_cmo_feature_init - FW_FEATURE_CMO is not stored in ibm,hypertas-functions,
* handle that here. (Stolen from parse_system_parameter_string)
*/
void pSeries_cmo_feature_init(void)
{
char *ptr, *key, *value, *end;
int call_status;
int page_order = IOMMU_PAGE_SHIFT;
pr_debug(" -> fw_cmo_feature_init()\n");
spin_lock(&rtas_data_buf_lock);
memset(rtas_data_buf, 0, RTAS_DATA_BUF_SIZE);
call_status = rtas_call(rtas_token("ibm,get-system-parameter"), 3, 1,
NULL,
CMO_CHARACTERISTICS_TOKEN,
__pa(rtas_data_buf),
RTAS_DATA_BUF_SIZE);
if (call_status != 0) {
spin_unlock(&rtas_data_buf_lock);
pr_debug("CMO not available\n");
pr_debug(" <- fw_cmo_feature_init()\n");
return;
}
end = rtas_data_buf + CMO_MAXLENGTH - 2;
ptr = rtas_data_buf + 2; /* step over strlen value */
key = value = ptr;
while (*ptr && (ptr <= end)) {
/* Separate the key and value by replacing '=' with '\0' and
* point the value at the string after the '='
*/
if (ptr[0] == '=') {
ptr[0] = '\0';
value = ptr + 1;
} else if (ptr[0] == '\0' || ptr[0] == ',') {
/* Terminate the string containing the key/value pair */
ptr[0] = '\0';
if (key == value) {
pr_debug("Malformed key/value pair\n");
/* Never found a '=', end processing */
break;
}
if (0 == strcmp(key, "CMOPageSize"))
page_order = simple_strtol(value, NULL, 10);
else if (0 == strcmp(key, "PrPSP"))
CMO_PrPSP = simple_strtol(value, NULL, 10);
else if (0 == strcmp(key, "SecPSP"))
CMO_SecPSP = simple_strtol(value, NULL, 10);
value = key = ptr + 1;
}
ptr++;
}
/* Page size is returned as the power of 2 of the page size,
* convert to the page size in bytes before returning
*/
CMO_PageSize = 1 << page_order;
pr_debug("CMO_PageSize = %lu\n", CMO_PageSize);
if (CMO_PrPSP != -1 || CMO_SecPSP != -1) {
pr_info("CMO enabled\n");
pr_debug("CMO enabled, PrPSP=%d, SecPSP=%d\n", CMO_PrPSP,
CMO_SecPSP);
powerpc_firmware_features |= FW_FEATURE_CMO;
pSeries_coalesce_init();
} else
pr_debug("CMO not enabled, PrPSP=%d, SecPSP=%d\n", CMO_PrPSP,
CMO_SecPSP);
spin_unlock(&rtas_data_buf_lock);
pr_debug(" <- fw_cmo_feature_init()\n");
}
/*
* Early initialization. Relocation is on but do not reference unbolted pages
*/
static void __init pSeries_init_early(void)
{
pr_debug(" -> pSeries_init_early()\n");
powerpc/pseries: Re-implement HVSI as part of hvc_vio On pseries machines, consoles are provided by the hypervisor using a low level get_chars/put_chars type interface. However, this is really just a transport to the service processor which implements them either as "raw" console (networked consoles, HMC, ...) or as "hvsi" serial ports. The later is a simple packet protocol on top of the raw character interface that is supposed to convey additional "serial port" style semantics. In practice however, all it does is provide a way to read the CD line and set/clear our DTR line, that's it. We currently implement the "raw" protocol as an hvc console backend (/dev/hvcN) and the "hvsi" protocol using a separate tty driver (/dev/hvsi0). However this is quite impractical. The arbitrary difference between the two type of devices has been a major source of user (and distro) confusion. Additionally, there's an additional mini -hvsi implementation in the pseries platform code for our low level debug console and early boot kernel messages, which means code duplication, though that low level variant is impractical as it's incapable of doing the initial protocol negociation to establish the link to the FSP. This essentially replaces the dedicated hvsi driver and the platform udbg code completely by extending the existing hvc_vio backend used in "raw" mode so that: - It now supports HVSI as well - We add support for hvc backend providing tiocm{get,set} - It also provides a udbg interface for early debug and boot console This is overall less code, though this will only be obvious once we remove the old "hvsi" driver, which is still available for now. When the old driver is enabled, the new code still kicks in for the low level udbg console, replacing the old mini implementation in the platform code, it just doesn't provide the higher level "hvc" interface. In addition to producing generally simler code, this has several benefits over our current situation: - The user/distro only has to deal with /dev/hvcN for the hypervisor console, avoiding all sort of confusion that has plagued us in the past - The tty, kernel and low level debug console all use the same code base which supports the full protocol establishment process, thus the console is now available much earlier than it used to be with the old HVSI driver. The kernel console works much earlier and udbg is available much earlier too. Hackers can enable a hard coded very-early debug console as well that works with HVSI (previously that was only supported for the "raw" mode). I've tried to keep the same semantics as hvsi relative to how I react to things like CD changes, with some subtle differences though: - I clear DTR on close if HUPCL is set - Current hvsi triggers a hangup if it detects a up->down transition on CD (you can still open a console with CD down). My new implementation triggers a hangup if the link to the FSP is severed, and severs it upon detecting a up->down transition on CD. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2011-05-12 05:46:38 +02:00
#ifdef CONFIG_HVC_CONSOLE
if (firmware_has_feature(FW_FEATURE_LPAR))
powerpc/pseries: Re-implement HVSI as part of hvc_vio On pseries machines, consoles are provided by the hypervisor using a low level get_chars/put_chars type interface. However, this is really just a transport to the service processor which implements them either as "raw" console (networked consoles, HMC, ...) or as "hvsi" serial ports. The later is a simple packet protocol on top of the raw character interface that is supposed to convey additional "serial port" style semantics. In practice however, all it does is provide a way to read the CD line and set/clear our DTR line, that's it. We currently implement the "raw" protocol as an hvc console backend (/dev/hvcN) and the "hvsi" protocol using a separate tty driver (/dev/hvsi0). However this is quite impractical. The arbitrary difference between the two type of devices has been a major source of user (and distro) confusion. Additionally, there's an additional mini -hvsi implementation in the pseries platform code for our low level debug console and early boot kernel messages, which means code duplication, though that low level variant is impractical as it's incapable of doing the initial protocol negociation to establish the link to the FSP. This essentially replaces the dedicated hvsi driver and the platform udbg code completely by extending the existing hvc_vio backend used in "raw" mode so that: - It now supports HVSI as well - We add support for hvc backend providing tiocm{get,set} - It also provides a udbg interface for early debug and boot console This is overall less code, though this will only be obvious once we remove the old "hvsi" driver, which is still available for now. When the old driver is enabled, the new code still kicks in for the low level udbg console, replacing the old mini implementation in the platform code, it just doesn't provide the higher level "hvc" interface. In addition to producing generally simler code, this has several benefits over our current situation: - The user/distro only has to deal with /dev/hvcN for the hypervisor console, avoiding all sort of confusion that has plagued us in the past - The tty, kernel and low level debug console all use the same code base which supports the full protocol establishment process, thus the console is now available much earlier than it used to be with the old HVSI driver. The kernel console works much earlier and udbg is available much earlier too. Hackers can enable a hard coded very-early debug console as well that works with HVSI (previously that was only supported for the "raw" mode). I've tried to keep the same semantics as hvsi relative to how I react to things like CD changes, with some subtle differences though: - I clear DTR on close if HUPCL is set - Current hvsi triggers a hangup if it detects a up->down transition on CD (you can still open a console with CD down). My new implementation triggers a hangup if the link to the FSP is severed, and severs it upon detecting a up->down transition on CD. Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
2011-05-12 05:46:38 +02:00
hvc_vio_init_early();
#endif
if (firmware_has_feature(FW_FEATURE_XDABR))
ppc_md.set_dabr = pseries_set_xdabr;
else if (firmware_has_feature(FW_FEATURE_DABR))
ppc_md.set_dabr = pseries_set_dabr;
pSeries_cmo_feature_init();
iommu_init_early_pSeries();
pr_debug(" <- pSeries_init_early()\n");
}
/*
* Called very early, MMU is off, device-tree isn't unflattened
*/
static int __init pSeries_probe_hypertas(unsigned long node,
const char *uname, int depth,
void *data)
{
const char *hypertas;
unsigned long len;
if (depth != 1 ||
(strcmp(uname, "rtas") != 0 && strcmp(uname, "rtas@0") != 0))
return 0;
hypertas = of_get_flat_dt_prop(node, "ibm,hypertas-functions", &len);
if (!hypertas)
return 1;
powerpc_firmware_features |= FW_FEATURE_LPAR;
fw_feature_init(hypertas, len);
return 1;
}
static int __init pSeries_probe(void)
{
unsigned long root = of_get_flat_dt_root();
char *dtype = of_get_flat_dt_prop(root, "device_type", NULL);
if (dtype == NULL)
return 0;
if (strcmp(dtype, "chrp"))
return 0;
/* Cell blades firmware claims to be chrp while it's not. Until this
* is fixed, we need to avoid those here.
*/
if (of_flat_dt_is_compatible(root, "IBM,CPBW-1.0") ||
of_flat_dt_is_compatible(root, "IBM,CBEA"))
return 0;
pr_debug("pSeries detected, looking for LPAR capability...\n");
/* Now try to figure out if we are running on LPAR */
of_scan_flat_dt(pSeries_probe_hypertas, NULL);
if (firmware_has_feature(FW_FEATURE_LPAR))
hpte_init_lpar();
else
hpte_init_native();
pr_debug("Machine is%s LPAR !\n",
(powerpc_firmware_features & FW_FEATURE_LPAR) ? "" : " not");
return 1;
}
ppc64: Set up PCI tree from Open Firmware device tree This adds code which gives us the option on ppc64 of instantiating the PCI tree (the tree of pci_bus and pci_dev structs) from the Open Firmware device tree rather than by probing PCI configuration space. The OF device tree has a node for each PCI device and bridge in the system, with properties that tell us what addresses the firmware has configured for them and other details. There are a couple of reasons why this is needed. First, on systems with a hypervisor, there is a PCI-PCI bridge per slot under the PCI host bridges. These PCI-PCI bridges have special isolation features for virtualization. We can't write to their config space, and we are not supposed to be reading their config space either. The firmware tells us about the address ranges that they pass in the OF device tree. Secondly, on powermacs, the interrupt controller is in a PCI device that may be behind a PCI-PCI bridge. If we happened to take an interrupt just at the point when the device or a bridge on the path to it was disabled for probing, we would crash when we try to access the interrupt controller. I have implemented a platform-specific function which is called for each PCI bridge (host or PCI-PCI) to say whether the code should look in the device tree or use normal PCI probing for the devices under that bridge. On pSeries machines we use the device tree if we're running under a hypervisor, otherwise we use normal probing. On powermacs we use normal probing for the AGP bridge, since the device for the AGP bridge itself isn't shown in the device tree (at least on my G5), and the device tree for everything else. This has been tested on a dual G5 powermac, a partition on a POWER5 machine (running under the hypervisor), and a legacy iSeries partition. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-09-12 09:17:36 +02:00
static int pSeries_pci_probe_mode(struct pci_bus *bus)
{
if (firmware_has_feature(FW_FEATURE_LPAR))
ppc64: Set up PCI tree from Open Firmware device tree This adds code which gives us the option on ppc64 of instantiating the PCI tree (the tree of pci_bus and pci_dev structs) from the Open Firmware device tree rather than by probing PCI configuration space. The OF device tree has a node for each PCI device and bridge in the system, with properties that tell us what addresses the firmware has configured for them and other details. There are a couple of reasons why this is needed. First, on systems with a hypervisor, there is a PCI-PCI bridge per slot under the PCI host bridges. These PCI-PCI bridges have special isolation features for virtualization. We can't write to their config space, and we are not supposed to be reading their config space either. The firmware tells us about the address ranges that they pass in the OF device tree. Secondly, on powermacs, the interrupt controller is in a PCI device that may be behind a PCI-PCI bridge. If we happened to take an interrupt just at the point when the device or a bridge on the path to it was disabled for probing, we would crash when we try to access the interrupt controller. I have implemented a platform-specific function which is called for each PCI bridge (host or PCI-PCI) to say whether the code should look in the device tree or use normal PCI probing for the devices under that bridge. On pSeries machines we use the device tree if we're running under a hypervisor, otherwise we use normal probing. On powermacs we use normal probing for the AGP bridge, since the device for the AGP bridge itself isn't shown in the device tree (at least on my G5), and the device tree for everything else. This has been tested on a dual G5 powermac, a partition on a POWER5 machine (running under the hypervisor), and a legacy iSeries partition. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-09-12 09:17:36 +02:00
return PCI_PROBE_DEVTREE;
return PCI_PROBE_NORMAL;
}
/**
* pSeries_power_off - tell firmware about how to power off the system.
*
* This function calls either the power-off rtas token in normal cases
* or the ibm,power-off-ups token (if present & requested) in case of
* a power failure. If power-off token is used, power on will only be
* possible with power button press. If ibm,power-off-ups token is used
* it will allow auto poweron after power is restored.
*/
static void pSeries_power_off(void)
{
int rc;
int rtas_poweroff_ups_token = rtas_token("ibm,power-off-ups");
if (rtas_flash_term_hook)
rtas_flash_term_hook(SYS_POWER_OFF);
if (rtas_poweron_auto == 0 ||
rtas_poweroff_ups_token == RTAS_UNKNOWN_SERVICE) {
rc = rtas_call(rtas_token("power-off"), 2, 1, NULL, -1, -1);
printk(KERN_INFO "RTAS power-off returned %d\n", rc);
} else {
rc = rtas_call(rtas_poweroff_ups_token, 0, 1, NULL);
printk(KERN_INFO "RTAS ibm,power-off-ups returned %d\n", rc);
}
for (;;);
}
#ifndef CONFIG_PCI
void pSeries_final_fixup(void) { }
#endif
define_machine(pseries) {
.name = "pSeries",
.probe = pSeries_probe,
.setup_arch = pSeries_setup_arch,
.init_early = pSeries_init_early,
.show_cpuinfo = pSeries_show_cpuinfo,
.log_error = pSeries_log_error,
.pcibios_fixup = pSeries_final_fixup,
ppc64: Set up PCI tree from Open Firmware device tree This adds code which gives us the option on ppc64 of instantiating the PCI tree (the tree of pci_bus and pci_dev structs) from the Open Firmware device tree rather than by probing PCI configuration space. The OF device tree has a node for each PCI device and bridge in the system, with properties that tell us what addresses the firmware has configured for them and other details. There are a couple of reasons why this is needed. First, on systems with a hypervisor, there is a PCI-PCI bridge per slot under the PCI host bridges. These PCI-PCI bridges have special isolation features for virtualization. We can't write to their config space, and we are not supposed to be reading their config space either. The firmware tells us about the address ranges that they pass in the OF device tree. Secondly, on powermacs, the interrupt controller is in a PCI device that may be behind a PCI-PCI bridge. If we happened to take an interrupt just at the point when the device or a bridge on the path to it was disabled for probing, we would crash when we try to access the interrupt controller. I have implemented a platform-specific function which is called for each PCI bridge (host or PCI-PCI) to say whether the code should look in the device tree or use normal PCI probing for the devices under that bridge. On pSeries machines we use the device tree if we're running under a hypervisor, otherwise we use normal probing. On powermacs we use normal probing for the AGP bridge, since the device for the AGP bridge itself isn't shown in the device tree (at least on my G5), and the device tree for everything else. This has been tested on a dual G5 powermac, a partition on a POWER5 machine (running under the hypervisor), and a legacy iSeries partition. Signed-off-by: Paul Mackerras <paulus@samba.org>
2005-09-12 09:17:36 +02:00
.pci_probe_mode = pSeries_pci_probe_mode,
.restart = rtas_restart,
.power_off = pSeries_power_off,
.halt = rtas_halt,
.panic = rtas_os_term,
.get_boot_time = rtas_get_boot_time,
.get_rtc_time = rtas_get_rtc_time,
.set_rtc_time = rtas_set_rtc_time,
.calibrate_decr = generic_calibrate_decr,
.progress = rtas_progress,
.system_reset_exception = pSeries_system_reset_exception,
.machine_check_exception = pSeries_machine_check_exception,
#ifdef CONFIG_KEXEC
.machine_kexec = pSeries_machine_kexec,
#endif
};