From 800252976be89611ef86d6d04442a821018ed949 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 14 Nov 2011 11:18:59 +0100 Subject: [S390] wire up process_vm syscalls Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/unistd.h | 4 +++- arch/s390/kernel/compat_wrapper.S | 20 ++++++++++++++++++++ arch/s390/kernel/syscalls.S | 2 ++ 3 files changed, 25 insertions(+), 1 deletion(-) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/unistd.h b/arch/s390/include/asm/unistd.h index 404bdb9671b4..58de4c91c333 100644 --- a/arch/s390/include/asm/unistd.h +++ b/arch/s390/include/asm/unistd.h @@ -277,7 +277,9 @@ #define __NR_clock_adjtime 337 #define __NR_syncfs 338 #define __NR_setns 339 -#define NR_syscalls 340 +#define __NR_process_vm_readv 340 +#define __NR_process_vm_writev 341 +#define NR_syscalls 342 /* * There are some system calls that are not present on 64 bit, some diff --git a/arch/s390/kernel/compat_wrapper.S b/arch/s390/kernel/compat_wrapper.S index 5006a1d9f5d0..18c51df9fe06 100644 --- a/arch/s390/kernel/compat_wrapper.S +++ b/arch/s390/kernel/compat_wrapper.S @@ -1627,3 +1627,23 @@ ENTRY(sys_setns_wrapper) lgfr %r2,%r2 # int lgfr %r3,%r3 # int jg sys_setns + +ENTRY(compat_sys_process_vm_readv_wrapper) + lgfr %r2,%r2 # compat_pid_t + llgtr %r3,%r3 # struct compat_iovec __user * + llgfr %r4,%r4 # unsigned long + llgtr %r5,%r5 # struct compat_iovec __user * + llgfr %r6,%r6 # unsigned long + llgf %r0,164(%r15) # unsigned long + stg %r0,160(%r15) + jg sys_process_vm_readv + +ENTRY(compat_sys_process_vm_writev_wrapper) + lgfr %r2,%r2 # compat_pid_t + llgtr %r3,%r3 # struct compat_iovec __user * + llgfr %r4,%r4 # unsigned long + llgtr %r5,%r5 # struct compat_iovec __user * + llgfr %r6,%r6 # unsigned long + llgf %r0,164(%r15) # unsigned long + stg %r0,160(%r15) + jg sys_process_vm_writev diff --git a/arch/s390/kernel/syscalls.S b/arch/s390/kernel/syscalls.S index 73eb08c874fb..bcab2f04ba58 100644 --- a/arch/s390/kernel/syscalls.S +++ b/arch/s390/kernel/syscalls.S @@ -348,3 +348,5 @@ SYSCALL(sys_open_by_handle_at,sys_open_by_handle_at,compat_sys_open_by_handle_at SYSCALL(sys_clock_adjtime,sys_clock_adjtime,compat_sys_clock_adjtime_wrapper) SYSCALL(sys_syncfs,sys_syncfs,sys_syncfs_wrapper) SYSCALL(sys_setns,sys_setns,sys_setns_wrapper) +SYSCALL(sys_process_vm_readv,sys_process_vm_readv,compat_sys_process_vm_readv_wrapper) /* 340 */ +SYSCALL(sys_process_vm_writev,sys_process_vm_writev,compat_sys_process_vm_writev_wrapper) -- cgit v1.2.3 From 09b538833b85521d937a06faf61e6a3273253cc0 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Mon, 14 Nov 2011 11:19:00 +0100 Subject: [S390] fix pgste update logic The pgste_update_all / pgste_update_young and pgste_set_pte need to check if the pte entry contains a valid page address before the storage key can be accessed. In addition pgste_set_pte needs to set the access key and fetch protection bit of the new pte entry, not the old entry. Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/pgtable.h | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index 34ede0ea85a9..524d23b8610c 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -593,6 +593,8 @@ static inline pgste_t pgste_update_all(pte_t *ptep, pgste_t pgste) unsigned long address, bits; unsigned char skey; + if (!pte_present(*ptep)) + return pgste; address = pte_val(*ptep) & PAGE_MASK; skey = page_get_storage_key(address); bits = skey & (_PAGE_CHANGED | _PAGE_REFERENCED); @@ -625,6 +627,8 @@ static inline pgste_t pgste_update_young(pte_t *ptep, pgste_t pgste) #ifdef CONFIG_PGSTE int young; + if (!pte_present(*ptep)) + return pgste; young = page_reset_referenced(pte_val(*ptep) & PAGE_MASK); /* Transfer page referenced bit to pte software bit (host view) */ if (young || (pgste_val(pgste) & RCP_HR_BIT)) @@ -638,13 +642,15 @@ static inline pgste_t pgste_update_young(pte_t *ptep, pgste_t pgste) } -static inline void pgste_set_pte(pte_t *ptep, pgste_t pgste) +static inline void pgste_set_pte(pte_t *ptep, pgste_t pgste, pte_t entry) { #ifdef CONFIG_PGSTE unsigned long address; unsigned long okey, nkey; - address = pte_val(*ptep) & PAGE_MASK; + if (!pte_present(entry)) + return; + address = pte_val(entry) & PAGE_MASK; okey = nkey = page_get_storage_key(address); nkey &= ~(_PAGE_ACC_BITS | _PAGE_FP_BIT); /* Set page access key and fetch protection bit from pgste */ @@ -712,7 +718,7 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr, if (mm_has_pgste(mm)) { pgste = pgste_get_lock(ptep); - pgste_set_pte(ptep, pgste); + pgste_set_pte(ptep, pgste, entry); *ptep = entry; pgste_set_unlock(ptep, pgste); } else -- cgit v1.2.3 From fa2fb2f4a599c402bb2670dde27867dbbb7e3c45 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 14 Nov 2011 11:19:01 +0100 Subject: [S390] pfault: ignore leftover completion interrupts Ignore completion interrupts if the initial interrupt hasn't been received and the addressed task is not running. This case can only happen if leftover (pending) completion interrupt gets delivered which wasn't removed with the PFAULT CANCEL operation during cpu hotplug. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/mm/fault.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c index 1766def5bc3f..a9a301866b3c 100644 --- a/arch/s390/mm/fault.c +++ b/arch/s390/mm/fault.c @@ -587,8 +587,13 @@ static void pfault_interrupt(unsigned int ext_int_code, } else { /* Completion interrupt was faster than initial * interrupt. Set pfault_wait to -1 so the initial - * interrupt doesn't put the task to sleep. */ - tsk->thread.pfault_wait = -1; + * interrupt doesn't put the task to sleep. + * If the task is not running, ignore the completion + * interrupt since it must be a leftover of a PFAULT + * CANCEL operation which didn't remove all pending + * completion interrupts. */ + if (tsk->state == TASK_RUNNING) + tsk->thread.pfault_wait = -1; } put_task_struct(tsk); } else { -- cgit v1.2.3 From 7a2512b744e72377c3fa5976f06a3f343e155d1f Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Mon, 14 Nov 2011 11:19:02 +0100 Subject: [S390] incorrect note program header 'readelf -n' on the s390 vmlinux file generates lots of warnings about corrupt notes. The reason is that the 'NOTE' program header has incorrect file and memory sizes. The problem is that the section following the NOTES section do not switch to a different phdr and they get added to the NOTE program section. Add a dummy entry to the linker script that switches to the data phdr before the start of the RODATA section. Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/vmlinux.lds.S | 2 ++ 1 file changed, 2 insertions(+) (limited to 'arch/s390') diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S index 56fe6bc81fee..e4c79ebb40e6 100644 --- a/arch/s390/kernel/vmlinux.lds.S +++ b/arch/s390/kernel/vmlinux.lds.S @@ -43,6 +43,8 @@ SECTIONS NOTES :text :note + .dummy : { *(.dummy) } :data + RODATA #ifdef CONFIG_SHARED_KERNEL -- cgit v1.2.3 From 96603b505cb6c54782a27599afef65cc108ef5f2 Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Mon, 14 Nov 2011 11:19:03 +0100 Subject: [S390] Kconfig: Select CONFIG_KEXEC for CONFIG_CRASH_DUMP The kdump infrastructure is built on top of kexec. Therefore CONFIG_KEXEC has to be enabled when CONFIG_CRASH_DUMP is selected. Signed-off-by: Michael Holzheu Signed-off-by: Martin Schwidefsky --- arch/s390/Kconfig | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/s390') diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig index a9fbd43395f7..373679b3744a 100644 --- a/arch/s390/Kconfig +++ b/arch/s390/Kconfig @@ -572,6 +572,7 @@ config KEXEC config CRASH_DUMP bool "kernel crash dumps" depends on 64BIT + select KEXEC help Generate crash dump after being started by kexec. Crash dump kernels are loaded in the main kernel with kexec-tools -- cgit v1.2.3 From 3f25dc4fcbc371f86a61a6af759003ebd4965908 Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Mon, 14 Nov 2011 11:19:05 +0100 Subject: [S390] zfcpdump: Do not initialize zfcpdump in kdump mode When the kernel is started in kdump mode, zfcpdump should not be initialized because both dump methods can't be used at the same time. Signed-off-by: Michael Holzheu Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/setup.c | 4 +++- drivers/s390/char/zcore.c | 2 ++ 2 files changed, 5 insertions(+), 1 deletion(-) (limited to 'arch/s390') diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 8ac6bfa2786c..e58a462949b1 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -211,6 +211,8 @@ static void __init setup_zfcpdump(unsigned int console_devno) if (ipl_info.type != IPL_TYPE_FCP_DUMP) return; + if (OLDMEM_BASE) + return; if (console_devno != -1) sprintf(str, " cio_ignore=all,!0.0.%04x,!0.0.%04x", ipl_info.data.fcp.dev_id.devno, console_devno); @@ -482,7 +484,7 @@ static void __init setup_memory_end(void) #ifdef CONFIG_ZFCPDUMP - if (ipl_info.type == IPL_TYPE_FCP_DUMP) { + if (ipl_info.type == IPL_TYPE_FCP_DUMP && !OLDMEM_BASE) { memory_end = ZFCPDUMP_HSA_SIZE; memory_end_set = 1; } diff --git a/drivers/s390/char/zcore.c b/drivers/s390/char/zcore.c index 43068fbd0baa..1b6d9247fdc7 100644 --- a/drivers/s390/char/zcore.c +++ b/drivers/s390/char/zcore.c @@ -641,6 +641,8 @@ static int __init zcore_init(void) if (ipl_info.type != IPL_TYPE_FCP_DUMP) return -ENODATA; + if (OLDMEM_BASE) + return -ENODATA; zcore_dbf = debug_register("zcore", 4, 1, 4 * sizeof(long)); debug_register_view(zcore_dbf, &debug_sprintf_view); -- cgit v1.2.3 From cfa1e7e1d49c6f5f0b00b2cb890b521e5c0dc7ea Mon Sep 17 00:00:00 2001 From: Jan Glauber Date: Mon, 14 Nov 2011 11:19:06 +0100 Subject: [S390] avoid STCKF if running in ESA mode In ESA mode STCKF is not defined even if the facility bit is enabled. To prevent an illegal operation we must also check if we run a 64 bit kernel. To make the check perform well add the STCKF bit to the machine flags. Signed-off-by: Jan Glauber Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/setup.h | 3 +++ arch/s390/include/asm/timex.h | 2 +- arch/s390/kernel/early.c | 2 ++ 3 files changed, 6 insertions(+), 1 deletion(-) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/setup.h b/arch/s390/include/asm/setup.h index 5a099714df04..097183c70407 100644 --- a/arch/s390/include/asm/setup.h +++ b/arch/s390/include/asm/setup.h @@ -82,6 +82,7 @@ extern unsigned int user_mode; #define MACHINE_FLAG_LPAR (1UL << 12) #define MACHINE_FLAG_SPP (1UL << 13) #define MACHINE_FLAG_TOPOLOGY (1UL << 14) +#define MACHINE_FLAG_STCKF (1UL << 15) #define MACHINE_IS_VM (S390_lowcore.machine_flags & MACHINE_FLAG_VM) #define MACHINE_IS_KVM (S390_lowcore.machine_flags & MACHINE_FLAG_KVM) @@ -100,6 +101,7 @@ extern unsigned int user_mode; #define MACHINE_HAS_PFMF (0) #define MACHINE_HAS_SPP (0) #define MACHINE_HAS_TOPOLOGY (0) +#define MACHINE_HAS_STCKF (0) #else /* __s390x__ */ #define MACHINE_HAS_IEEE (1) #define MACHINE_HAS_CSP (1) @@ -111,6 +113,7 @@ extern unsigned int user_mode; #define MACHINE_HAS_PFMF (S390_lowcore.machine_flags & MACHINE_FLAG_PFMF) #define MACHINE_HAS_SPP (S390_lowcore.machine_flags & MACHINE_FLAG_SPP) #define MACHINE_HAS_TOPOLOGY (S390_lowcore.machine_flags & MACHINE_FLAG_TOPOLOGY) +#define MACHINE_HAS_STCKF (S390_lowcore.machine_flags & MACHINE_FLAG_STCKF) #endif /* __s390x__ */ #define ZFCPDUMP_HSA_SIZE (32UL<<20) diff --git a/arch/s390/include/asm/timex.h b/arch/s390/include/asm/timex.h index d610bef9c5e9..c447a27a7fdb 100644 --- a/arch/s390/include/asm/timex.h +++ b/arch/s390/include/asm/timex.h @@ -90,7 +90,7 @@ static inline unsigned long long get_clock_fast(void) { unsigned long long clk; - if (test_facility(25)) + if (MACHINE_HAS_STCKF) asm volatile(".insn s,0xb27c0000,%0" : "=Q" (clk) : : "cc"); else clk = get_clock(); diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 37394b3413e2..c9ffe0025197 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -390,6 +390,8 @@ static __init void detect_machine_facilities(void) S390_lowcore.machine_flags |= MACHINE_FLAG_MVCOS; if (test_facility(40)) S390_lowcore.machine_flags |= MACHINE_FLAG_SPP; + if (test_facility(25)) + S390_lowcore.machine_flags |= MACHINE_FLAG_STCKF; #endif } -- cgit v1.2.3 From 6ed54387dc470fc439cb154724a1ac81d251c126 Mon Sep 17 00:00:00 2001 From: Jan Glauber Date: Mon, 14 Nov 2011 11:19:07 +0100 Subject: [S390] crypto: avoid MSA3 and MSA4 instructions in ESA mode MSA3 and MSA4 instructions are only available under CONFIG_64BIT. Bail out before using any of these instructions if the kernel is running in 31 bit mode. Signed-off-by: Jan Glauber Signed-off-by: Martin Schwidefsky --- arch/s390/crypto/crypt_s390.h | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/crypto/crypt_s390.h b/arch/s390/crypto/crypt_s390.h index 49676771bd66..ffd1ac255f19 100644 --- a/arch/s390/crypto/crypt_s390.h +++ b/arch/s390/crypto/crypt_s390.h @@ -368,9 +368,12 @@ static inline int crypt_s390_func_available(int func, if (facility_mask & CRYPT_S390_MSA && !test_facility(17)) return 0; - if (facility_mask & CRYPT_S390_MSA3 && !test_facility(76)) + + if (facility_mask & CRYPT_S390_MSA3 && + (!test_facility(2) || !test_facility(76))) return 0; - if (facility_mask & CRYPT_S390_MSA4 && !test_facility(77)) + if (facility_mask & CRYPT_S390_MSA4 && + (!test_facility(2) || !test_facility(77))) return 0; switch (func & CRYPT_S390_OP_MASK) { -- cgit v1.2.3 From f6bf1a8acd2cb3a92a7b7c9ab03e56a32ac5ece5 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 14 Nov 2011 11:19:08 +0100 Subject: [S390] topology: fix topology on z10 machines Make sure that all cpus in a book on a z10 appear as book siblings and not as core siblings. This fixes some performance regressions that appeared after the book scheduling domain got introduced. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/topology.c | 45 ++++++++++++++++++++++++++++++++++++++------- 1 file changed, 38 insertions(+), 7 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c index 77b8942b9a15..fdb5b8cb260f 100644 --- a/arch/s390/kernel/topology.c +++ b/arch/s390/kernel/topology.c @@ -68,8 +68,10 @@ static cpumask_t cpu_group_map(struct mask_info *info, unsigned int cpu) return mask; } -static void add_cpus_to_mask(struct topology_cpu *tl_cpu, - struct mask_info *book, struct mask_info *core) +static struct mask_info *add_cpus_to_mask(struct topology_cpu *tl_cpu, + struct mask_info *book, + struct mask_info *core, + int z10) { unsigned int cpu; @@ -88,10 +90,16 @@ static void add_cpus_to_mask(struct topology_cpu *tl_cpu, cpu_book_id[lcpu] = book->id; #endif cpumask_set_cpu(lcpu, &core->mask); - cpu_core_id[lcpu] = core->id; + if (z10) { + cpu_core_id[lcpu] = rcpu; + core = core->next; + } else { + cpu_core_id[lcpu] = core->id; + } smp_cpu_polarization[lcpu] = tl_cpu->pp; } } + return core; } static void clear_masks(void) @@ -123,18 +131,41 @@ static void tl_to_cores(struct sysinfo_15_1_x *info) { #ifdef CONFIG_SCHED_BOOK struct mask_info *book = &book_info; + struct cpuid cpu_id; #else struct mask_info *book = NULL; #endif struct mask_info *core = &core_info; union topology_entry *tle, *end; + int z10 = 0; - +#ifdef CONFIG_SCHED_BOOK + get_cpu_id(&cpu_id); + z10 = cpu_id.machine == 0x2097 || cpu_id.machine == 0x2098; +#endif spin_lock_irq(&topology_lock); clear_masks(); tle = info->tle; end = (union topology_entry *)((unsigned long)info + info->length); while (tle < end) { +#ifdef CONFIG_SCHED_BOOK + if (z10) { + switch (tle->nl) { + case 1: + book = book->next; + book->id = tle->container.id; + break; + case 0: + core = add_cpus_to_mask(&tle->cpu, book, core, z10); + break; + default: + clear_masks(); + goto out; + } + tle = next_tle(tle); + continue; + } +#endif switch (tle->nl) { #ifdef CONFIG_SCHED_BOOK case 2: @@ -147,7 +178,7 @@ static void tl_to_cores(struct sysinfo_15_1_x *info) core->id = tle->container.id; break; case 0: - add_cpus_to_mask(&tle->cpu, book, core); + add_cpus_to_mask(&tle->cpu, book, core, z10); break; default: clear_masks(); @@ -328,8 +359,8 @@ void __init s390_init_cpu_topology(void) for (i = 0; i < TOPOLOGY_NR_MAG; i++) printk(" %d", info->mag[i]); printk(" / %d\n", info->mnest); - alloc_masks(info, &core_info, 2); + alloc_masks(info, &core_info, 1); #ifdef CONFIG_SCHED_BOOK - alloc_masks(info, &book_info, 3); + alloc_masks(info, &book_info, 2); #endif } -- cgit v1.2.3 From 9e6dabeffd1d0ec2aa19aa076c4886067238d442 Mon Sep 17 00:00:00 2001 From: Cornelia Huck Date: Thu, 17 Nov 2011 11:00:41 +0100 Subject: KVM: s390: Fix RUNNING flag misinterpretation CPUSTAT_RUNNING was implemented signifying that a vcpu is not stopped. This is not, however, what the architecture says: RUNNING should be set when the host is acting on the behalf of the guest operating system. CPUSTAT_RUNNING has been changed to be set in kvm_arch_vcpu_load() and to be unset in kvm_arch_vcpu_put(). For signifying stopped state of a vcpu, a host-controlled bit has been used and is set/unset basically on the reverse as the old CPUSTAT_RUNNING bit (including pushing it down into stop handling proper in handle_stop()). Cc: stable@kernel.org Signed-off-by: Cornelia Huck Signed-off-by: Carsten Otte Signed-off-by: Avi Kivity --- arch/s390/include/asm/kvm_host.h | 2 +- arch/s390/kvm/diag.c | 2 +- arch/s390/kvm/intercept.c | 3 ++- arch/s390/kvm/interrupt.c | 1 + arch/s390/kvm/kvm-s390.c | 10 +++++++--- arch/s390/kvm/sigp.c | 6 +++--- 6 files changed, 15 insertions(+), 9 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h index 24e18473d926..20f9068cd6b5 100644 --- a/arch/s390/include/asm/kvm_host.h +++ b/arch/s390/include/asm/kvm_host.h @@ -47,7 +47,7 @@ struct sca_block { #define KVM_HPAGE_MASK(x) (~(KVM_HPAGE_SIZE(x) - 1)) #define KVM_PAGES_PER_HPAGE(x) (KVM_HPAGE_SIZE(x) / PAGE_SIZE) -#define CPUSTAT_HOST 0x80000000 +#define CPUSTAT_STOPPED 0x80000000 #define CPUSTAT_WAIT 0x10000000 #define CPUSTAT_ECALL_PEND 0x08000000 #define CPUSTAT_STOP_INT 0x04000000 diff --git a/arch/s390/kvm/diag.c b/arch/s390/kvm/diag.c index 87cedd61be04..8943e82cd4d9 100644 --- a/arch/s390/kvm/diag.c +++ b/arch/s390/kvm/diag.c @@ -70,7 +70,7 @@ static int __diag_ipl_functions(struct kvm_vcpu *vcpu) return -EOPNOTSUPP; } - atomic_clear_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); + atomic_set_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags); vcpu->run->s390_reset_flags |= KVM_S390_RESET_SUBSYSTEM; vcpu->run->s390_reset_flags |= KVM_S390_RESET_IPL; vcpu->run->s390_reset_flags |= KVM_S390_RESET_CPU_INIT; diff --git a/arch/s390/kvm/intercept.c b/arch/s390/kvm/intercept.c index c7c51898984e..02434543eabb 100644 --- a/arch/s390/kvm/intercept.c +++ b/arch/s390/kvm/intercept.c @@ -132,7 +132,6 @@ static int handle_stop(struct kvm_vcpu *vcpu) int rc = 0; vcpu->stat.exit_stop_request++; - atomic_clear_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); spin_lock_bh(&vcpu->arch.local_int.lock); if (vcpu->arch.local_int.action_bits & ACTION_STORE_ON_STOP) { vcpu->arch.local_int.action_bits &= ~ACTION_STORE_ON_STOP; @@ -149,6 +148,8 @@ static int handle_stop(struct kvm_vcpu *vcpu) } if (vcpu->arch.local_int.action_bits & ACTION_STOP_ON_STOP) { + atomic_set_mask(CPUSTAT_STOPPED, + &vcpu->arch.sie_block->cpuflags); vcpu->arch.local_int.action_bits &= ~ACTION_STOP_ON_STOP; VCPU_EVENT(vcpu, 3, "%s", "cpu stopped"); rc = -EOPNOTSUPP; diff --git a/arch/s390/kvm/interrupt.c b/arch/s390/kvm/interrupt.c index 87c16705b381..278ee009ce65 100644 --- a/arch/s390/kvm/interrupt.c +++ b/arch/s390/kvm/interrupt.c @@ -252,6 +252,7 @@ static void __do_deliver_interrupt(struct kvm_vcpu *vcpu, offsetof(struct _lowcore, restart_psw), sizeof(psw_t)); if (rc == -EFAULT) exception = 1; + atomic_clear_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags); break; case KVM_S390_PROGRAM_INT: diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index 0bd3bea1e4cd..630ab65b488b 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -270,10 +270,12 @@ void kvm_arch_vcpu_load(struct kvm_vcpu *vcpu, int cpu) restore_fp_regs(&vcpu->arch.guest_fpregs); restore_access_regs(vcpu->arch.guest_acrs); gmap_enable(vcpu->arch.gmap); + atomic_set_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); } void kvm_arch_vcpu_put(struct kvm_vcpu *vcpu) { + atomic_clear_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); gmap_disable(vcpu->arch.gmap); save_fp_regs(&vcpu->arch.guest_fpregs); save_access_regs(vcpu->arch.guest_acrs); @@ -301,7 +303,9 @@ static void kvm_s390_vcpu_initial_reset(struct kvm_vcpu *vcpu) int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu) { - atomic_set(&vcpu->arch.sie_block->cpuflags, CPUSTAT_ZARCH | CPUSTAT_SM); + atomic_set(&vcpu->arch.sie_block->cpuflags, CPUSTAT_ZARCH | + CPUSTAT_SM | + CPUSTAT_STOPPED); vcpu->arch.sie_block->ecb = 6; vcpu->arch.sie_block->eca = 0xC1002001U; vcpu->arch.sie_block->fac = (int) (long) facilities; @@ -428,7 +432,7 @@ static int kvm_arch_vcpu_ioctl_set_initial_psw(struct kvm_vcpu *vcpu, psw_t psw) { int rc = 0; - if (atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_RUNNING) + if (!(atomic_read(&vcpu->arch.sie_block->cpuflags) & CPUSTAT_STOPPED)) rc = -EBUSY; else { vcpu->run->psw_mask = psw.mask; @@ -501,7 +505,7 @@ rerun_vcpu: if (vcpu->sigset_active) sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved); - atomic_set_mask(CPUSTAT_RUNNING, &vcpu->arch.sie_block->cpuflags); + atomic_clear_mask(CPUSTAT_STOPPED, &vcpu->arch.sie_block->cpuflags); BUG_ON(vcpu->kvm->arch.float_int.local_int[vcpu->vcpu_id] == NULL); diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c index f815118835f3..4a488a98d7f8 100644 --- a/arch/s390/kvm/sigp.c +++ b/arch/s390/kvm/sigp.c @@ -57,8 +57,8 @@ static int __sigp_sense(struct kvm_vcpu *vcpu, u16 cpu_addr, spin_lock(&fi->lock); if (fi->local_int[cpu_addr] == NULL) rc = 3; /* not operational */ - else if (atomic_read(fi->local_int[cpu_addr]->cpuflags) - & CPUSTAT_RUNNING) { + else if (!(atomic_read(fi->local_int[cpu_addr]->cpuflags) + & CPUSTAT_STOPPED)) { *reg &= 0xffffffff00000000UL; rc = 1; /* status stored */ } else { @@ -251,7 +251,7 @@ static int __sigp_set_prefix(struct kvm_vcpu *vcpu, u16 cpu_addr, u32 address, spin_lock_bh(&li->lock); /* cpu must be in stopped state */ - if (atomic_read(li->cpuflags) & CPUSTAT_RUNNING) { + if (!(atomic_read(li->cpuflags) & CPUSTAT_STOPPED)) { rc = 1; /* incorrect state */ *reg &= SIGP_STAT_INCORRECT_STATE; kfree(inti); -- cgit v1.2.3 From bd59d3a4444e5f941c863db1a7715edf7289e29a Mon Sep 17 00:00:00 2001 From: Cornelia Huck Date: Thu, 17 Nov 2011 11:00:42 +0100 Subject: KVM: s390: handle SIGP sense running intercepts SIGP sense running may cause an intercept on higher level virtualization, so handle it by checking the CPUSTAT_RUNNING flag. Signed-off-by: Cornelia Huck Signed-off-by: Carsten Otte Signed-off-by: Avi Kivity --- arch/s390/include/asm/kvm_host.h | 1 + arch/s390/kvm/kvm-s390.c | 1 + arch/s390/kvm/sigp.c | 39 +++++++++++++++++++++++++++++++++++++++ 3 files changed, 41 insertions(+) (limited to 'arch/s390') diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h index 20f9068cd6b5..b0c235cb6ad5 100644 --- a/arch/s390/include/asm/kvm_host.h +++ b/arch/s390/include/asm/kvm_host.h @@ -139,6 +139,7 @@ struct kvm_vcpu_stat { u32 instruction_stfl; u32 instruction_tprot; u32 instruction_sigp_sense; + u32 instruction_sigp_sense_running; u32 instruction_sigp_external_call; u32 instruction_sigp_emergency; u32 instruction_sigp_stop; diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index 630ab65b488b..0891926ea953 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -65,6 +65,7 @@ struct kvm_stats_debugfs_item debugfs_entries[] = { { "instruction_stfl", VCPU_STAT(instruction_stfl) }, { "instruction_tprot", VCPU_STAT(instruction_tprot) }, { "instruction_sigp_sense", VCPU_STAT(instruction_sigp_sense) }, + { "instruction_sigp_sense_running", VCPU_STAT(instruction_sigp_sense_running) }, { "instruction_sigp_external_call", VCPU_STAT(instruction_sigp_external_call) }, { "instruction_sigp_emergency", VCPU_STAT(instruction_sigp_emergency) }, { "instruction_sigp_stop", VCPU_STAT(instruction_sigp_stop) }, diff --git a/arch/s390/kvm/sigp.c b/arch/s390/kvm/sigp.c index 4a488a98d7f8..0a7941d74bc6 100644 --- a/arch/s390/kvm/sigp.c +++ b/arch/s390/kvm/sigp.c @@ -31,9 +31,11 @@ #define SIGP_SET_PREFIX 0x0d #define SIGP_STORE_STATUS_ADDR 0x0e #define SIGP_SET_ARCH 0x12 +#define SIGP_SENSE_RUNNING 0x15 /* cpu status bits */ #define SIGP_STAT_EQUIPMENT_CHECK 0x80000000UL +#define SIGP_STAT_NOT_RUNNING 0x00000400UL #define SIGP_STAT_INCORRECT_STATE 0x00000200UL #define SIGP_STAT_INVALID_PARAMETER 0x00000100UL #define SIGP_STAT_EXT_CALL_PENDING 0x00000080UL @@ -275,6 +277,38 @@ out_fi: return rc; } +static int __sigp_sense_running(struct kvm_vcpu *vcpu, u16 cpu_addr, + unsigned long *reg) +{ + int rc; + struct kvm_s390_float_interrupt *fi = &vcpu->kvm->arch.float_int; + + if (cpu_addr >= KVM_MAX_VCPUS) + return 3; /* not operational */ + + spin_lock(&fi->lock); + if (fi->local_int[cpu_addr] == NULL) + rc = 3; /* not operational */ + else { + if (atomic_read(fi->local_int[cpu_addr]->cpuflags) + & CPUSTAT_RUNNING) { + /* running */ + rc = 1; + } else { + /* not running */ + *reg &= 0xffffffff00000000UL; + *reg |= SIGP_STAT_NOT_RUNNING; + rc = 0; + } + } + spin_unlock(&fi->lock); + + VCPU_EVENT(vcpu, 4, "sensed running status of cpu %x rc %x", cpu_addr, + rc); + + return rc; +} + int kvm_s390_handle_sigp(struct kvm_vcpu *vcpu) { int r1 = (vcpu->arch.sie_block->ipa & 0x00f0) >> 4; @@ -331,6 +365,11 @@ int kvm_s390_handle_sigp(struct kvm_vcpu *vcpu) rc = __sigp_set_prefix(vcpu, cpu_addr, parameter, &vcpu->arch.guest_gprs[r1]); break; + case SIGP_SENSE_RUNNING: + vcpu->stat.instruction_sigp_sense_running++; + rc = __sigp_sense_running(vcpu, cpu_addr, + &vcpu->arch.guest_gprs[r1]); + break; case SIGP_RESTART: vcpu->stat.instruction_sigp_restart++; /* user space must know about restart */ -- cgit v1.2.3 From 1eddb85f889a5a9fe59ec184e06844716a00aded Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Thu, 17 Nov 2011 11:00:43 +0100 Subject: KVM: s390: Fix tprot locking There is a potential host deadlock in the tprot intercept handling. We must not hold the mmap semaphore while resolving the guest address. If userspace is remapping, then the memory detection in the guest is broken anyway so we can safely separate the address translation from walking the vmas. Signed-off-by: Christian Borntraeger Signed-off-by: Carsten Otte Signed-off-by: Avi Kivity --- arch/s390/kvm/priv.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'arch/s390') diff --git a/arch/s390/kvm/priv.c b/arch/s390/kvm/priv.c index 391626361084..d02638959922 100644 --- a/arch/s390/kvm/priv.c +++ b/arch/s390/kvm/priv.c @@ -336,6 +336,7 @@ static int handle_tprot(struct kvm_vcpu *vcpu) u64 address1 = disp1 + base1 ? vcpu->arch.guest_gprs[base1] : 0; u64 address2 = disp2 + base2 ? vcpu->arch.guest_gprs[base2] : 0; struct vm_area_struct *vma; + unsigned long user_address; vcpu->stat.instruction_tprot++; @@ -349,9 +350,14 @@ static int handle_tprot(struct kvm_vcpu *vcpu) return -EOPNOTSUPP; + /* we must resolve the address without holding the mmap semaphore. + * This is ok since the userspace hypervisor is not supposed to change + * the mapping while the guest queries the memory. Otherwise the guest + * might crash or get wrong info anyway. */ + user_address = (unsigned long) __guestaddr_to_user(vcpu, address1); + down_read(¤t->mm->mmap_sem); - vma = find_vma(current->mm, - (unsigned long) __guestaddr_to_user(vcpu, address1)); + vma = find_vma(current->mm, user_address); if (!vma) { up_read(¤t->mm->mmap_sem); return kvm_s390_inject_program_int(vcpu, PGM_ADDRESSING); -- cgit v1.2.3 From 52e16b185fdd0aeb0522c203e944873e38ac1575 Mon Sep 17 00:00:00 2001 From: Christian Borntraeger Date: Thu, 17 Nov 2011 11:00:44 +0100 Subject: KVM: s390: announce SYNC_MMU KVM on s390 always had a sync mmu. Any mapping change in userspace mapping was always reflected immediately in the guest mapping. - In older code the guest mapping was just an offset - In newer code the last level page table is shared Signed-off-by: Christian Borntraeger Signed-off-by: Carsten Otte Signed-off-by: Avi Kivity --- arch/s390/kvm/kvm-s390.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/s390') diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index 0891926ea953..d1c445732451 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -128,6 +128,7 @@ int kvm_dev_ioctl_check_extension(long ext) switch (ext) { case KVM_CAP_S390_PSW: case KVM_CAP_S390_GMAP: + case KVM_CAP_SYNC_MMU: r = 1; break; default: -- cgit v1.2.3