Merge branch 'sthyi' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux into kvms390/next

get sthyi rework with the KVM changes

13 files changed, 182 insertions, 71 deletions

diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index 51375e766e90..fd006a272024 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -736,7 +736,6 @@ struct kvm_arch{
         wait_queue_head_t ipte_wq;
         int ipte_lock_count;
         struct mutex ipte_mutex;
-        struct ratelimit_state sthyi_limit;
         spinlock_t start_stop_lock;
         struct sie_page2 *sie_page2;
         struct kvm_s390_cpu_model model;
diff --git a/arch/s390/include/asm/sysinfo.h b/arch/s390/include/asm/sysinfo.h
index 2b498e58b914..e4a28307bc5d 100644
--- a/arch/s390/include/asm/sysinfo.h
+++ b/arch/s390/include/asm/sysinfo.h
@@ -198,4 +198,5 @@ struct service_level {
 int register_service_level(struct service_level *);
 int unregister_service_level(struct service_level *);
 
+int sthyi_fill(void *dst, u64 *rc);
 #endif /* __ASM_S390_SYSINFO_H */
diff --git a/arch/s390/include/uapi/asm/sthyi.h b/arch/s390/include/uapi/asm/sthyi.h
new file mode 100644
index 000000000000..ec113db4eb7e
--- /dev/null
+++ b/arch/s390/include/uapi/asm/sthyi.h
@@ -0,0 +1,6 @@
+#ifndef _UAPI_ASM_STHYI_H
+#define _UAPI_ASM_STHYI_H
+
+#define STHYI_FC_CP_IFL_CAP     0
+
+#endif /* _UAPI_ASM_STHYI_H */
diff --git a/arch/s390/include/uapi/asm/unistd.h b/arch/s390/include/uapi/asm/unistd.h
index ea42290e7d51..61c64f543769 100644
--- a/arch/s390/include/uapi/asm/unistd.h
+++ b/arch/s390/include/uapi/asm/unistd.h
@@ -315,7 +315,8 @@
 #define __NR_pwritev2           377
 #define __NR_s390_guarded_storage       378
 #define __NR_statx              379
-#define NR_syscalls 380
+#define __NR_s390_sthyi         380
+#define NR_syscalls 381
 
 /* 
  * There are some system calls that are not present on 64 bit, some
diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile
index adb3fe2e3d42..1fefb7f9216f 100644
--- a/arch/s390/kernel/Makefile
+++ b/arch/s390/kernel/Makefile
@@ -55,7 +55,7 @@ obj-y	:= traps.o time.o process.o base.o early.o setup.o idle.o vtime.o
 obj-y   += processor.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o
 obj-y   += debug.o irq.o ipl.o dis.o diag.o vdso.o als.o
 obj-y   += sysinfo.o jump_label.o lgr.o os_info.o machine_kexec.o pgm_check.o
-obj-y   += runtime_instr.o cache.o fpu.o dumpstack.o guarded_storage.o
+obj-y   += runtime_instr.o cache.o fpu.o dumpstack.o guarded_storage.o sthyi.o
 obj-y   += entry.o reipl.o relocate_kernel.o kdebugfs.o
 
 extra-y                         += head.o head64.o vmlinux.lds
diff --git a/arch/s390/kernel/compat_wrapper.c b/arch/s390/kernel/compat_wrapper.c
index 986642a3543b..eb0b17ed95b6 100644
--- a/arch/s390/kernel/compat_wrapper.c
+++ b/arch/s390/kernel/compat_wrapper.c
@@ -180,3 +180,4 @@ COMPAT_SYSCALL_WRAP3(mlock2, unsigned long, start, size_t, len, int, flags);
 COMPAT_SYSCALL_WRAP6(copy_file_range, int, fd_in, loff_t __user *, off_in, int, fd_out, loff_t __user *, off_out, size_t, len, unsigned int, flags);
 COMPAT_SYSCALL_WRAP2(s390_guarded_storage, int, command, struct gs_cb *, gs_cb);
 COMPAT_SYSCALL_WRAP5(statx, int, dfd, const char __user *, path, unsigned, flags, unsigned, mask, struct statx __user *, buffer);
+COMPAT_SYSCALL_WRAP4(s390_sthyi, unsigned long, code, void __user *, info, u64 __user *, rc, unsigned long, flags);
diff --git a/arch/s390/kernel/entry.h b/arch/s390/kernel/entry.h
index dbf5f7e18246..bb5301eeb4f4 100644
--- a/arch/s390/kernel/entry.h
+++ b/arch/s390/kernel/entry.h
@@ -77,6 +77,7 @@ long sys_s390_runtime_instr(int command, int signum);
 long sys_s390_guarded_storage(int command, struct gs_cb __user *);
 long sys_s390_pci_mmio_write(unsigned long, const void __user *, size_t);
 long sys_s390_pci_mmio_read(unsigned long, void __user *, size_t);
+long sys_s390_sthyi(unsigned long function_code, void __user *buffer, u64 __user *return_code, unsigned long flags);
 
 DECLARE_PER_CPU(u64, mt_cycles[8]);
 
diff --git a/arch/s390/kvm/sthyi.c b/arch/s390/kernel/sthyi.c
index 395926b8c1ed..12981e197f01 100644
--- a/arch/s390/kvm/sthyi.c
+++ b/arch/s390/kernel/sthyi.c
@@ -8,22 +8,19 @@
  * Copyright IBM Corp. 2016
  * Author(s): Janosch Frank <frankja@linux.vnet.ibm.com>
  */
-#include <linux/kvm_host.h>
 #include <linux/errno.h>
 #include <linux/pagemap.h>
 #include <linux/vmalloc.h>
-#include <linux/ratelimit.h>
-
-#include <asm/kvm_host.h>
+#include <linux/syscalls.h>
+#include <linux/mutex.h>
 #include <asm/asm-offsets.h>
 #include <asm/sclp.h>
 #include <asm/diag.h>
 #include <asm/sysinfo.h>
 #include <asm/ebcdic.h>
-
-#include "kvm-s390.h"
-#include "gaccess.h"
-#include "trace.h"
+#include <asm/facility.h>
+#include <asm/sthyi.h>
+#include "entry.h"
 
 #define DED_WEIGHT 0xffff
 /*
@@ -144,6 +141,21 @@ struct lpar_cpu_inf {
         struct cpu_inf ifl;
 };
 
+/*
+ * STHYI requires extensive locking in the higher hypervisors
+ * and is very computational/memory expensive. Therefore we
+ * cache the retrieved data whose valid period is 1s.
+ */
+#define CACHE_VALID_JIFFIES     HZ
+
+struct sthyi_info {
+        void *info;
+        unsigned long end;
+};
+
+static DEFINE_MUTEX(sthyi_mutex);
+static struct sthyi_info sthyi_cache;
+
 static inline u64 cpu_id(u8 ctidx, void *diag224_buf)
 {
         return *((u64 *)(diag224_buf + (ctidx + 1) * DIAG204_CPU_NAME_LEN));
@@ -382,88 +394,124 @@ out:
         vfree(diag204_buf);
 }
 
-static int sthyi(u64 vaddr)
+static int sthyi(u64 vaddr, u64 *rc)
 {
         register u64 code asm("0") = 0;
         register u64 addr asm("2") = vaddr;
+        register u64 rcode asm("3");
         int cc;
 
         asm volatile(
                 ".insn   rre,0xB2560000,%[code],%[addr]\n"
                 "ipm     %[cc]\n"
                 "srl     %[cc],28\n"
-                : [cc] "=d" (cc)
+                : [cc] "=d" (cc), "=d" (rcode)
                 : [code] "d" (code), [addr] "a" (addr)
-                : "3", "memory", "cc");
+                : "memory", "cc");
+        *rc = rcode;
         return cc;
 }
 
-int handle_sthyi(struct kvm_vcpu *vcpu)
+static int fill_dst(void *dst, u64 *rc)
 {
-        int reg1, reg2, r = 0;
-        u64 code, addr, cc = 0;
-        struct sthyi_sctns *sctns = NULL;
-
-        if (!test_kvm_facility(vcpu->kvm, 74))
-                return kvm_s390_inject_program_int(vcpu, PGM_OPERATION);
+        struct sthyi_sctns *sctns = (struct sthyi_sctns *)dst;
 
         /*
-         * STHYI requires extensive locking in the higher hypervisors
-         * and is very computational/memory expensive. Therefore we
-         * ratelimit the executions per VM.
+         * If the facility is on, we don't want to emulate the instruction.
+         * We ask the hypervisor to provide the data.
          */
-        if (!__ratelimit(&vcpu->kvm->arch.sthyi_limit)) {
-                kvm_s390_retry_instr(vcpu);
+        if (test_facility(74))
+                return sthyi((u64)dst, rc);
+
+        fill_hdr(sctns);
+        fill_stsi(sctns);
+        fill_diag(sctns);
+        *rc = 0;
+        return 0;
+}
+
+static int sthyi_init_cache(void)
+{
+        if (sthyi_cache.info)
                 return 0;
-        }
+        sthyi_cache.info = (void *)get_zeroed_page(GFP_KERNEL);
+        if (!sthyi_cache.info)
+                return -ENOMEM;
+        sthyi_cache.end = jiffies - 1; /* expired */
+        return 0;
+}
 
-        kvm_s390_get_regs_rre(vcpu, &reg1, &reg2);
-        code = vcpu->run->s.regs.gprs[reg1];
-        addr = vcpu->run->s.regs.gprs[reg2];
+static int sthyi_update_cache(u64 *rc)
+{
+        int r;
 
-        vcpu->stat.instruction_sthyi++;
-        VCPU_EVENT(vcpu, 3, "STHYI: fc: %llu addr: 0x%016llx", code, addr);
-        trace_kvm_s390_handle_sthyi(vcpu, code, addr);
+        memset(sthyi_cache.info, 0, PAGE_SIZE);
+        r = fill_dst(sthyi_cache.info, rc);
+        if (r)
+                return r;
+        sthyi_cache.end = jiffies + CACHE_VALID_JIFFIES;
+        return r;
+}
 
-        if (reg1 == reg2 || reg1 & 1 || reg2 & 1)
-                return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
+/*
+ * sthyi_fill - Fill page with data returned by the STHYI instruction
+ *
+ * @dst: Pointer to zeroed page
+ * @rc:  Pointer for storing the return code of the instruction
+ *
+ * Fills the destination with system information returned by the STHYI
+ * instruction. The data is generated by emulation or execution of STHYI,
+ * if available. The return value is the condition code that would be
+ * returned, the rc parameter is the return code which is passed in
+ * register R2 + 1.
+ */
+int sthyi_fill(void *dst, u64 *rc)
+{
+        int r;
 
-        if (code & 0xffff) {
-                cc = 3;
+        mutex_lock(&sthyi_mutex);
+        r = sthyi_init_cache();
+        if (r)
                 goto out;
-        }
 
-        if (addr & ~PAGE_MASK)
-                return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
+        if (time_is_before_jiffies(sthyi_cache.end)) {
+                /* cache expired */
+                r = sthyi_update_cache(rc);
+                if (r)
+                        goto out;
+        }
+        *rc = 0;
+        memcpy(dst, sthyi_cache.info, PAGE_SIZE);
+out:
+        mutex_unlock(&sthyi_mutex);
+        return r;
+}
+EXPORT_SYMBOL_GPL(sthyi_fill);
 
-        sctns = (void *)get_zeroed_page(GFP_KERNEL);
-        if (!sctns)
+SYSCALL_DEFINE4(s390_sthyi, unsigned long, function_code, void __user *, buffer,
+                u64 __user *, return_code, unsigned long, flags)
+{
+        u64 sthyi_rc;
+        void *info;
+        int r;
+
+        if (flags)
+                return -EINVAL;
+        if (function_code != STHYI_FC_CP_IFL_CAP)
+                return -EOPNOTSUPP;
+        info = (void *)get_zeroed_page(GFP_KERNEL);
+        if (!info)
                 return -ENOMEM;
-
-        /*
-         * If we are a guest, we don't want to emulate an emulated
-         * instruction. We ask the hypervisor to provide the data.
-         */
-        if (test_facility(74)) {
-                cc = sthyi((u64)sctns);
+        r = sthyi_fill(info, &sthyi_rc);
+        if (r < 0)
+                goto out;
+        if (return_code && put_user(sthyi_rc, return_code)) {
+                r = -EFAULT;
                 goto out;
         }
-
-        fill_hdr(sctns);
-        fill_stsi(sctns);
-        fill_diag(sctns);
-
+        if (copy_to_user(buffer, info, PAGE_SIZE))
+                r = -EFAULT;
 out:
-        if (!cc) {
-                r = write_guest(vcpu, addr, reg2, sctns, PAGE_SIZE);
-                if (r) {
-                        free_page((unsigned long)sctns);
-                        return kvm_s390_inject_prog_cond(vcpu, r);
-                }
-        }
-
-        free_page((unsigned long)sctns);
-        vcpu->run->s.regs.gprs[reg2 + 1] = cc ? 4 : 0;
-        kvm_s390_set_psw_cc(vcpu, cc);
+        free_page((unsigned long)info);
         return r;
 }
diff --git a/arch/s390/kernel/syscalls.S b/arch/s390/kernel/syscalls.S
index 54fce7b065de..0fb407ebbf46 100644
--- a/arch/s390/kernel/syscalls.S
+++ b/arch/s390/kernel/syscalls.S
@@ -388,3 +388,4 @@ SYSCALL(sys_preadv2,compat_sys_preadv2)
 SYSCALL(sys_pwritev2,compat_sys_pwritev2)
 SYSCALL(sys_s390_guarded_storage,compat_sys_s390_guarded_storage) /* 378 */
 SYSCALL(sys_statx,compat_sys_statx)
+SYSCALL(sys_s390_sthyi,compat_sys_s390_sthyi)
diff --git a/arch/s390/kvm/Makefile b/arch/s390/kvm/Makefile
index 09a9e6dfc09f..6048b1c6e580 100644
--- a/arch/s390/kvm/Makefile
+++ b/arch/s390/kvm/Makefile
@@ -12,6 +12,6 @@ common-objs = $(KVM)/kvm_main.o $(KVM)/eventfd.o  $(KVM)/async_pf.o $(KVM)/irqch
 ccflags-y := -Ivirt/kvm -Iarch/s390/kvm
 
 kvm-objs := $(common-objs) kvm-s390.o intercept.o interrupt.o priv.o sigp.o
-kvm-objs += diag.o gaccess.o guestdbg.o sthyi.o vsie.o
+kvm-objs += diag.o gaccess.o guestdbg.o vsie.o
 
 obj-$(CONFIG_KVM) += kvm.o
diff --git a/arch/s390/kvm/intercept.c b/arch/s390/kvm/intercept.c
index a4752bf6b526..8fe034beb623 100644
--- a/arch/s390/kvm/intercept.c
+++ b/arch/s390/kvm/intercept.c
@@ -18,6 +18,7 @@
 #include <asm/kvm_host.h>
 #include <asm/asm-offsets.h>
 #include <asm/irq.h>
+#include <asm/sysinfo.h>
 
 #include "kvm-s390.h"
 #include "gaccess.h"
@@ -360,6 +361,61 @@ static int handle_partial_execution(struct kvm_vcpu *vcpu)
         return -EOPNOTSUPP;
 }
 
+/*
+ * Handle the sthyi instruction that provides the guest with system
+ * information, like current CPU resources available at each level of
+ * the machine.
+ */
+int handle_sthyi(struct kvm_vcpu *vcpu)
+{
+        int reg1, reg2, r = 0;
+        u64 code, addr, cc = 0, rc = 0;
+        struct sthyi_sctns *sctns = NULL;
+
+        if (!test_kvm_facility(vcpu->kvm, 74))
+                return kvm_s390_inject_program_int(vcpu, PGM_OPERATION);
+
+        kvm_s390_get_regs_rre(vcpu, &reg1, &reg2);
+        code = vcpu->run->s.regs.gprs[reg1];
+        addr = vcpu->run->s.regs.gprs[reg2];
+
+        vcpu->stat.instruction_sthyi++;
+        VCPU_EVENT(vcpu, 3, "STHYI: fc: %llu addr: 0x%016llx", code, addr);
+        trace_kvm_s390_handle_sthyi(vcpu, code, addr);
+
+        if (reg1 == reg2 || reg1 & 1 || reg2 & 1)
+                return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
+
+        if (code & 0xffff) {
+                cc = 3;
+                rc = 4;
+                goto out;
+        }
+
+        if (addr & ~PAGE_MASK)
+                return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
+
+        sctns = (void *)get_zeroed_page(GFP_KERNEL);
+        if (!sctns)
+                return -ENOMEM;
+
+        cc = sthyi_fill(sctns, &rc);
+
+out:
+        if (!cc) {
+                r = write_guest(vcpu, addr, reg2, sctns, PAGE_SIZE);
+                if (r) {
+                        free_page((unsigned long)sctns);
+                        return kvm_s390_inject_prog_cond(vcpu, r);
+                }
+        }
+
+        free_page((unsigned long)sctns);
+        vcpu->run->s.regs.gprs[reg2 + 1] = rc;
+        kvm_s390_set_psw_cc(vcpu, cc);
+        return r;
+}
+
 static int handle_operexc(struct kvm_vcpu *vcpu)
 {
         psw_t oldpsw, newpsw;
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 40d0a1a97889..de6a5b790da0 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -1884,8 +1884,6 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 
         rc = -ENOMEM;
 
-        ratelimit_state_init(&kvm->arch.sthyi_limit, 5 * HZ, 500);
-
         kvm->arch.use_esca = 0; /* start with basic SCA */
         if (!sclp.has_64bscao)
                 alloc_flags |= GFP_DMA;
diff --git a/arch/s390/kvm/kvm-s390.h b/arch/s390/kvm/kvm-s390.h
index 9f8fdd7b2311..10d65dfbc306 100644
--- a/arch/s390/kvm/kvm-s390.h
+++ b/arch/s390/kvm/kvm-s390.h
@@ -242,6 +242,8 @@ static inline void kvm_s390_retry_instr(struct kvm_vcpu *vcpu)
         kvm_s390_rewind_psw(vcpu, kvm_s390_get_ilen(vcpu));
 }
 
+int handle_sthyi(struct kvm_vcpu *vcpu);
+
 /* implemented in priv.c */
 int is_valid_psw(psw_t *psw);
 int kvm_s390_handle_aa(struct kvm_vcpu *vcpu);
@@ -268,9 +270,6 @@ void kvm_s390_vsie_destroy(struct kvm *kvm);
 int kvm_s390_handle_sigp(struct kvm_vcpu *vcpu);
 int kvm_s390_handle_sigp_pei(struct kvm_vcpu *vcpu);
 
-/* implemented in sthyi.c */
-int handle_sthyi(struct kvm_vcpu *vcpu);
-
 /* implemented in kvm-s390.c */
 void kvm_s390_set_tod_clock_ext(struct kvm *kvm,
                                  const struct kvm_s390_vm_tod_clock *gtod);