* [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-30 18:18 ` Aurelien Jarno
2009-11-26 13:23 ` [Qemu-devel] [PATCH 02/11] S/390 host/target build system support Alexander Graf
` (10 subsequent siblings)
11 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
Because Qemu currently requires a TCG target to exist and there are quite some
useful helpers here to lay the groundwork for out KVM target, let's create a
stub TCG emulation target for S390X CPUs.
This is required to make tcg happy. The emulation target itself won't work
though.
Signed-off-by: Alexander Graf <agraf@suse.de>
---
cpu-exec.c | 2 +
target-s390x/cpu.h | 119 ++++++++++++++++++++++++++++++++++++++++++++++
target-s390x/exec.h | 51 ++++++++++++++++++++
target-s390x/helper.c | 57 ++++++++++++++++++++++
target-s390x/op_helper.c | 74 ++++++++++++++++++++++++++++
target-s390x/translate.c | 57 ++++++++++++++++++++++
6 files changed, 360 insertions(+), 0 deletions(-)
create mode 100644 target-s390x/cpu.h
create mode 100644 target-s390x/exec.h
create mode 100644 target-s390x/helper.c
create mode 100644 target-s390x/op_helper.c
create mode 100644 target-s390x/translate.c
diff --git a/cpu-exec.c b/cpu-exec.c
index 2c0765c..af4595b 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -249,6 +249,7 @@ int cpu_exec(CPUState *env1)
#elif defined(TARGET_MIPS)
#elif defined(TARGET_SH4)
#elif defined(TARGET_CRIS)
+#elif defined(TARGET_S390X)
/* XXXXX */
#else
#error unsupported target CPU
@@ -673,6 +674,7 @@ int cpu_exec(CPUState *env1)
#elif defined(TARGET_SH4)
#elif defined(TARGET_ALPHA)
#elif defined(TARGET_CRIS)
+#elif defined(TARGET_S390X)
/* XXXXX */
#else
#error unsupported target CPU
diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
new file mode 100644
index 0000000..f45b00c
--- /dev/null
+++ b/target-s390x/cpu.h
@@ -0,0 +1,119 @@
+/*
+ * S/390 virtual CPU header
+ *
+ * Copyright (c) 2009 Ulrich Hecht
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
+ */
+#ifndef CPU_S390X_H
+#define CPU_S390X_H
+
+#define TARGET_LONG_BITS 64
+
+#define ELF_MACHINE EM_S390
+
+#define CPUState struct CPUS390XState
+
+#include "cpu-defs.h"
+
+#include "softfloat.h"
+
+#define NB_MMU_MODES 2 // guess
+#define MMU_USER_IDX 0 // guess
+
+typedef union FPReg {
+ struct {
+#ifdef WORDS_BIGENDIAN
+ float32 e;
+ int32_t __pad;
+#else
+ int32_t __pad;
+ float32 e;
+#endif
+ };
+ float64 d;
+ uint64_t i;
+} FPReg;
+
+typedef struct CPUS390XState {
+ uint64_t regs[16]; /* GP registers */
+
+ uint32_t aregs[16]; /* access registers */
+
+ uint32_t fpc; /* floating-point control register */
+ FPReg fregs[16]; /* FP registers */
+ float_status fpu_status; /* passed to softfloat lib */
+
+ struct {
+ uint64_t mask;
+ uint64_t addr;
+ } psw;
+
+ int cc; /* condition code (0-3) */
+
+ uint64_t __excp_addr;
+
+ CPU_COMMON
+} CPUS390XState;
+
+#if defined(CONFIG_USER_ONLY)
+static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
+{
+ if (newsp)
+ env->regs[15] = newsp;
+ env->regs[0] = 0;
+}
+#endif
+
+CPUS390XState *cpu_s390x_init(const char *cpu_model);
+int cpu_s390x_exec(CPUS390XState *s);
+void cpu_s390x_close(CPUS390XState *s);
+
+/* you can call this signal handler from your SIGBUS and SIGSEGV
+ signal handlers to inform the virtual CPU of exceptions. non zero
+ is returned if the signal was handled by the virtual CPU. */
+int cpu_s390x_signal_handler(int host_signum, void *pinfo,
+ void *puc);
+int cpu_s390x_handle_mmu_fault (CPUS390XState *env, target_ulong address, int rw,
+ int mmu_idx, int is_softmuu);
+#define cpu_handle_mmu_fault cpu_s390x_handle_mmu_fault
+
+#define TARGET_PAGE_BITS 12
+
+#define cpu_init cpu_s390x_init
+#define cpu_exec cpu_s390x_exec
+#define cpu_gen_code cpu_s390x_gen_code
+
+#include "cpu-all.h"
+#include "exec-all.h"
+
+#define EXCP_OPEX 1 /* operation exception (sigill) */
+#define EXCP_SVC 2 /* supervisor call (syscall) */
+#define EXCP_ADDR 5 /* addressing exception */
+#define EXCP_EXECUTE_SVC 0xff00000 /* supervisor call via execute insn */
+
+static inline void cpu_pc_from_tb(CPUState *env, TranslationBlock* tb)
+{
+ env->psw.addr = tb->pc;
+}
+
+static inline void cpu_get_tb_cpu_state(CPUState* env, target_ulong *pc,
+ target_ulong *cs_base, int *flags)
+{
+ *pc = env->psw.addr;
+ *cs_base = 0;
+ *flags = env->psw.mask; // guess
+}
+#endif
diff --git a/target-s390x/exec.h b/target-s390x/exec.h
new file mode 100644
index 0000000..5198359
--- /dev/null
+++ b/target-s390x/exec.h
@@ -0,0 +1,51 @@
+/*
+ * S/390 execution defines
+ *
+ * Copyright (c) 2009 Ulrich Hecht
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
+ */
+
+#include "dyngen-exec.h"
+
+register struct CPUS390XState *env asm(AREG0);
+
+#include "cpu.h"
+#include "exec-all.h"
+
+static inline int cpu_has_work(CPUState *env)
+{
+ return env->interrupt_request & CPU_INTERRUPT_HARD; // guess
+}
+
+static inline void regs_to_env(void)
+{
+}
+
+static inline void env_to_regs(void)
+{
+}
+
+static inline int cpu_halted(CPUState *env)
+{
+ if (!env->halted) {
+ return 0;
+ }
+ if (cpu_has_work(env)) {
+ env->halted = 0;
+ return 0;
+ }
+ return EXCP_HALTED;
+}
diff --git a/target-s390x/helper.c b/target-s390x/helper.c
new file mode 100644
index 0000000..4e23b4a
--- /dev/null
+++ b/target-s390x/helper.c
@@ -0,0 +1,57 @@
+/*
+ * S/390 helpers
+ *
+ * Copyright (c) 2009 Ulrich Hecht
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+
+#include "cpu.h"
+#include "exec-all.h"
+#include "gdbstub.h"
+#include "qemu-common.h"
+
+CPUS390XState *cpu_s390x_init(const char *cpu_model)
+{
+ CPUS390XState *env;
+ static int inited = 0;
+
+ env = qemu_mallocz(sizeof(CPUS390XState));
+ cpu_exec_init(env);
+ if (!inited) {
+ inited = 1;
+ }
+
+ env->cpu_model_str = cpu_model;
+ cpu_reset(env);
+ qemu_init_vcpu(env);
+ return env;
+}
+
+void cpu_reset(CPUS390XState *env)
+{
+ if (qemu_loglevel_mask(CPU_LOG_RESET)) {
+ qemu_log("CPU Reset (CPU %d)\n", env->cpu_index);
+ log_cpu_state(env, 0);
+ }
+
+ memset(env, 0, offsetof(CPUS390XState, breakpoints));
+ /* FIXME: reset vector? */
+ tlb_flush(env, 1);
+}
diff --git a/target-s390x/op_helper.c b/target-s390x/op_helper.c
new file mode 100644
index 0000000..e623de9
--- /dev/null
+++ b/target-s390x/op_helper.c
@@ -0,0 +1,74 @@
+/*
+ * S/390 helper routines
+ *
+ * Copyright (c) 2009 Alexander Graf
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
+ */
+
+#include "exec.h"
+
+/*****************************************************************************/
+/* Softmmu support */
+#if !defined (CONFIG_USER_ONLY)
+
+#define MMUSUFFIX _mmu
+
+#define SHIFT 0
+#include "softmmu_template.h"
+
+#define SHIFT 1
+#include "softmmu_template.h"
+
+#define SHIFT 2
+#include "softmmu_template.h"
+
+#define SHIFT 3
+#include "softmmu_template.h"
+
+/* try to fill the TLB and return an exception if error. If retaddr is
+ NULL, it means that the function was called in C code (i.e. not
+ from generated code or from helper.c) */
+/* XXX: fix it to restore all registers */
+void tlb_fill (target_ulong addr, int is_write, int mmu_idx, void *retaddr)
+{
+ TranslationBlock *tb;
+ CPUState *saved_env;
+ unsigned long pc;
+ int ret;
+
+ /* XXX: hack to restore env in all cases, even if not called from
+ generated code */
+ saved_env = env;
+ env = cpu_single_env;
+ ret = cpu_s390x_handle_mmu_fault(env, addr, is_write, mmu_idx, 1);
+ if (unlikely(ret != 0)) {
+ if (likely(retaddr)) {
+ /* now we have a real cpu fault */
+ pc = (unsigned long)retaddr;
+ tb = tb_find_pc(pc);
+ if (likely(tb)) {
+ /* the PC is inside the translated code. It means that we have
+ a virtual CPU fault */
+ cpu_restore_state(tb, env, pc, NULL);
+ }
+ }
+ /* XXX */
+ /* helper_raise_exception_err(env->exception_index, env->error_code); */
+ }
+ env = saved_env;
+}
+
+#endif
diff --git a/target-s390x/translate.c b/target-s390x/translate.c
new file mode 100644
index 0000000..f304411
--- /dev/null
+++ b/target-s390x/translate.c
@@ -0,0 +1,57 @@
+/*
+ * S/390 translation
+ *
+ * Copyright (c) 2009 Ulrich Hecht
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
+ */
+
+#include "cpu.h"
+#include "exec-all.h"
+#include "disas.h"
+#include "tcg-op.h"
+#include "qemu-log.h"
+
+void cpu_dump_state(CPUState *env, FILE *f,
+ int (*cpu_fprintf)(FILE *f, const char *fmt, ...),
+ int flags)
+{
+ int i;
+ for (i = 0; i < 16; i++) {
+ cpu_fprintf(f, "R%02d=%016lx", i, env->regs[i]);
+ if ((i % 4) == 3) cpu_fprintf(f, "\n");
+ else cpu_fprintf(f, " ");
+ }
+ for (i = 0; i < 16; i++) {
+ cpu_fprintf(f, "F%02d=%016lx", i, env->fregs[i]);
+ if ((i % 4) == 3) cpu_fprintf(f, "\n");
+ else cpu_fprintf(f, " ");
+ }
+ cpu_fprintf(f, "PSW=mask %016lx addr %016lx cc %02x\n", env->psw.mask, env->psw.addr, env->cc);
+}
+
+void gen_intermediate_code (CPUState *env, struct TranslationBlock *tb)
+{
+}
+
+void gen_intermediate_code_pc (CPUState *env, struct TranslationBlock *tb)
+{
+}
+
+void gen_pc_load(CPUState *env, TranslationBlock *tb,
+ unsigned long searched_pc, int pc_pos, void *puc)
+{
+ env->psw.addr = gen_opc_pc[pc_pos];
+}
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-11-26 13:23 ` [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation Alexander Graf
@ 2009-11-30 18:18 ` Aurelien Jarno
2009-11-30 22:30 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-11-30 18:18 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Thu, Nov 26, 2009 at 02:23:10PM +0100, Alexander Graf wrote:
> Because Qemu currently requires a TCG target to exist and there are quite some
> useful helpers here to lay the groundwork for out KVM target, let's create a
> stub TCG emulation target for S390X CPUs.
>
> This is required to make tcg happy. The emulation target itself won't work
> though.
Please find the comments below.
> Signed-off-by: Alexander Graf <agraf@suse.de>
> ---
> cpu-exec.c | 2 +
> target-s390x/cpu.h | 119 ++++++++++++++++++++++++++++++++++++++++++++++
> target-s390x/exec.h | 51 ++++++++++++++++++++
> target-s390x/helper.c | 57 ++++++++++++++++++++++
> target-s390x/op_helper.c | 74 ++++++++++++++++++++++++++++
> target-s390x/translate.c | 57 ++++++++++++++++++++++
> 6 files changed, 360 insertions(+), 0 deletions(-)
> create mode 100644 target-s390x/cpu.h
> create mode 100644 target-s390x/exec.h
> create mode 100644 target-s390x/helper.c
> create mode 100644 target-s390x/op_helper.c
> create mode 100644 target-s390x/translate.c
>
> diff --git a/cpu-exec.c b/cpu-exec.c
> index 2c0765c..af4595b 100644
> --- a/cpu-exec.c
> +++ b/cpu-exec.c
> @@ -249,6 +249,7 @@ int cpu_exec(CPUState *env1)
> #elif defined(TARGET_MIPS)
> #elif defined(TARGET_SH4)
> #elif defined(TARGET_CRIS)
> +#elif defined(TARGET_S390X)
> /* XXXXX */
> #else
> #error unsupported target CPU
> @@ -673,6 +674,7 @@ int cpu_exec(CPUState *env1)
> #elif defined(TARGET_SH4)
> #elif defined(TARGET_ALPHA)
> #elif defined(TARGET_CRIS)
> +#elif defined(TARGET_S390X)
> /* XXXXX */
> #else
> #error unsupported target CPU
> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
> new file mode 100644
> index 0000000..f45b00c
> --- /dev/null
> +++ b/target-s390x/cpu.h
> @@ -0,0 +1,119 @@
> +/*
> + * S/390 virtual CPU header
> + *
> + * Copyright (c) 2009 Ulrich Hecht
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
> + */
> +#ifndef CPU_S390X_H
> +#define CPU_S390X_H
> +
> +#define TARGET_LONG_BITS 64
> +
> +#define ELF_MACHINE EM_S390
> +
> +#define CPUState struct CPUS390XState
> +
> +#include "cpu-defs.h"
> +
> +#include "softfloat.h"
> +
> +#define NB_MMU_MODES 2 // guess
> +#define MMU_USER_IDX 0 // guess
> +
> +typedef union FPReg {
> + struct {
> +#ifdef WORDS_BIGENDIAN
> + float32 e;
> + int32_t __pad;
> +#else
> + int32_t __pad;
> + float32 e;
> +#endif
> + };
> + float64 d;
> + uint64_t i;
> +} FPReg;
> +
> +typedef struct CPUS390XState {
> + uint64_t regs[16]; /* GP registers */
> +
> + uint32_t aregs[16]; /* access registers */
> +
> + uint32_t fpc; /* floating-point control register */
> + FPReg fregs[16]; /* FP registers */
> + float_status fpu_status; /* passed to softfloat lib */
> +
> + struct {
> + uint64_t mask;
> + uint64_t addr;
> + } psw;
> +
> + int cc; /* condition code (0-3) */
> +
> + uint64_t __excp_addr;
> +
> + CPU_COMMON
> +} CPUS390XState;
> +
> +#if defined(CONFIG_USER_ONLY)
> +static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
> +{
> + if (newsp)
> + env->regs[15] = newsp;
Coding style.
> + env->regs[0] = 0;
> +}
> +#endif
> +
> +CPUS390XState *cpu_s390x_init(const char *cpu_model);
> +int cpu_s390x_exec(CPUS390XState *s);
> +void cpu_s390x_close(CPUS390XState *s);
> +
> +/* you can call this signal handler from your SIGBUS and SIGSEGV
> + signal handlers to inform the virtual CPU of exceptions. non zero
> + is returned if the signal was handled by the virtual CPU. */
> +int cpu_s390x_signal_handler(int host_signum, void *pinfo,
> + void *puc);
> +int cpu_s390x_handle_mmu_fault (CPUS390XState *env, target_ulong address, int rw,
> + int mmu_idx, int is_softmuu);
> +#define cpu_handle_mmu_fault cpu_s390x_handle_mmu_fault
> +
> +#define TARGET_PAGE_BITS 12
> +
> +#define cpu_init cpu_s390x_init
> +#define cpu_exec cpu_s390x_exec
> +#define cpu_gen_code cpu_s390x_gen_code
> +
> +#include "cpu-all.h"
> +#include "exec-all.h"
> +
> +#define EXCP_OPEX 1 /* operation exception (sigill) */
> +#define EXCP_SVC 2 /* supervisor call (syscall) */
> +#define EXCP_ADDR 5 /* addressing exception */
> +#define EXCP_EXECUTE_SVC 0xff00000 /* supervisor call via execute insn */
> +
> +static inline void cpu_pc_from_tb(CPUState *env, TranslationBlock* tb)
> +{
> + env->psw.addr = tb->pc;
> +}
> +
> +static inline void cpu_get_tb_cpu_state(CPUState* env, target_ulong *pc,
> + target_ulong *cs_base, int *flags)
> +{
> + *pc = env->psw.addr;
> + *cs_base = 0;
> + *flags = env->psw.mask; // guess
I don't know what psw.mask represent, but it may be wrong. It should be
a way to identify which TB can be reused, that is they have been
generated in the same CPU mode.
> +}
> +#endif
> diff --git a/target-s390x/exec.h b/target-s390x/exec.h
> new file mode 100644
> index 0000000..5198359
> --- /dev/null
> +++ b/target-s390x/exec.h
> @@ -0,0 +1,51 @@
> +/*
> + * S/390 execution defines
> + *
> + * Copyright (c) 2009 Ulrich Hecht
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
> + */
> +
> +#include "dyngen-exec.h"
> +
> +register struct CPUS390XState *env asm(AREG0);
> +
> +#include "cpu.h"
> +#include "exec-all.h"
> +
> +static inline int cpu_has_work(CPUState *env)
> +{
> + return env->interrupt_request & CPU_INTERRUPT_HARD; // guess
> +}
> +
> +static inline void regs_to_env(void)
> +{
> +}
> +
> +static inline void env_to_regs(void)
> +{
> +}
> +
> +static inline int cpu_halted(CPUState *env)
> +{
> + if (!env->halted) {
> + return 0;
> + }
> + if (cpu_has_work(env)) {
> + env->halted = 0;
> + return 0;
> + }
> + return EXCP_HALTED;
> +}
> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
> new file mode 100644
> index 0000000..4e23b4a
> --- /dev/null
> +++ b/target-s390x/helper.c
> @@ -0,0 +1,57 @@
> +/*
> + * S/390 helpers
> + *
> + * Copyright (c) 2009 Ulrich Hecht
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
> + */
> +
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <string.h>
> +
> +#include "cpu.h"
> +#include "exec-all.h"
> +#include "gdbstub.h"
> +#include "qemu-common.h"
> +
> +CPUS390XState *cpu_s390x_init(const char *cpu_model)
> +{
> + CPUS390XState *env;
> + static int inited = 0;
> +
> + env = qemu_mallocz(sizeof(CPUS390XState));
> + cpu_exec_init(env);
> + if (!inited) {
> + inited = 1;
> + }
> +
> + env->cpu_model_str = cpu_model;
> + cpu_reset(env);
> + qemu_init_vcpu(env);
> + return env;
> +}
> +
> +void cpu_reset(CPUS390XState *env)
> +{
> + if (qemu_loglevel_mask(CPU_LOG_RESET)) {
> + qemu_log("CPU Reset (CPU %d)\n", env->cpu_index);
> + log_cpu_state(env, 0);
> + }
> +
> + memset(env, 0, offsetof(CPUS390XState, breakpoints));
> + /* FIXME: reset vector? */
Isn't it also needed by kvm to make reset working?
> + tlb_flush(env, 1);
> +}
> diff --git a/target-s390x/op_helper.c b/target-s390x/op_helper.c
> new file mode 100644
> index 0000000..e623de9
> --- /dev/null
> +++ b/target-s390x/op_helper.c
> @@ -0,0 +1,74 @@
> +/*
> + * S/390 helper routines
> + *
> + * Copyright (c) 2009 Alexander Graf
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
> + */
> +
> +#include "exec.h"
> +
> +/*****************************************************************************/
> +/* Softmmu support */
> +#if !defined (CONFIG_USER_ONLY)
> +
> +#define MMUSUFFIX _mmu
> +
> +#define SHIFT 0
> +#include "softmmu_template.h"
> +
> +#define SHIFT 1
> +#include "softmmu_template.h"
> +
> +#define SHIFT 2
> +#include "softmmu_template.h"
> +
> +#define SHIFT 3
> +#include "softmmu_template.h"
> +
> +/* try to fill the TLB and return an exception if error. If retaddr is
> + NULL, it means that the function was called in C code (i.e. not
> + from generated code or from helper.c) */
> +/* XXX: fix it to restore all registers */
> +void tlb_fill (target_ulong addr, int is_write, int mmu_idx, void *retaddr)
> +{
> + TranslationBlock *tb;
> + CPUState *saved_env;
> + unsigned long pc;
> + int ret;
> +
> + /* XXX: hack to restore env in all cases, even if not called from
> + generated code */
> + saved_env = env;
> + env = cpu_single_env;
> + ret = cpu_s390x_handle_mmu_fault(env, addr, is_write, mmu_idx, 1);
> + if (unlikely(ret != 0)) {
> + if (likely(retaddr)) {
> + /* now we have a real cpu fault */
> + pc = (unsigned long)retaddr;
> + tb = tb_find_pc(pc);
> + if (likely(tb)) {
> + /* the PC is inside the translated code. It means that we have
> + a virtual CPU fault */
> + cpu_restore_state(tb, env, pc, NULL);
> + }
> + }
> + /* XXX */
> + /* helper_raise_exception_err(env->exception_index, env->error_code); */
> + }
> + env = saved_env;
> +}
> +
> +#endif
> diff --git a/target-s390x/translate.c b/target-s390x/translate.c
> new file mode 100644
> index 0000000..f304411
> --- /dev/null
> +++ b/target-s390x/translate.c
> @@ -0,0 +1,57 @@
> +/*
> + * S/390 translation
> + *
> + * Copyright (c) 2009 Ulrich Hecht
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, write to the Free Software
> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
> + */
> +
> +#include "cpu.h"
> +#include "exec-all.h"
> +#include "disas.h"
> +#include "tcg-op.h"
> +#include "qemu-log.h"
> +
> +void cpu_dump_state(CPUState *env, FILE *f,
> + int (*cpu_fprintf)(FILE *f, const char *fmt, ...),
> + int flags)
> +{
> + int i;
> + for (i = 0; i < 16; i++) {
> + cpu_fprintf(f, "R%02d=%016lx", i, env->regs[i]);
> + if ((i % 4) == 3) cpu_fprintf(f, "\n");
> + else cpu_fprintf(f, " ");
> + }
> + for (i = 0; i < 16; i++) {
> + cpu_fprintf(f, "F%02d=%016lx", i, env->fregs[i]);
> + if ((i % 4) == 3) cpu_fprintf(f, "\n");
> + else cpu_fprintf(f, " ");
> + }
> + cpu_fprintf(f, "PSW=mask %016lx addr %016lx cc %02x\n", env->psw.mask, env->psw.addr, env->cc);
> +}
> +
> +void gen_intermediate_code (CPUState *env, struct TranslationBlock *tb)
> +{
> +}
> +
> +void gen_intermediate_code_pc (CPUState *env, struct TranslationBlock *tb)
> +{
> +}
> +
> +void gen_pc_load(CPUState *env, TranslationBlock *tb,
> + unsigned long searched_pc, int pc_pos, void *puc)
> +{
> + env->psw.addr = gen_opc_pc[pc_pos];
> +}
> --
> 1.6.0.2
>
>
>
>
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-11-30 18:18 ` Aurelien Jarno
@ 2009-11-30 22:30 ` Alexander Graf
2009-12-01 9:46 ` Carsten Otte
2009-12-02 8:17 ` Aurelien Jarno
0 siblings, 2 replies; 49+ messages in thread
From: Alexander Graf @ 2009-11-30 22:30 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> On Thu, Nov 26, 2009 at 02:23:10PM +0100, Alexander Graf wrote:
>> Because Qemu currently requires a TCG target to exist and there are quite some
>> useful helpers here to lay the groundwork for out KVM target, let's create a
>> stub TCG emulation target for S390X CPUs.
>>
>> This is required to make tcg happy. The emulation target itself won't work
>> though.
>
> Please find the comments below.
>
>> Signed-off-by: Alexander Graf <agraf@suse.de>
>> ---
>> cpu-exec.c | 2 +
>> target-s390x/cpu.h | 119 ++++++++++++++++++++++++++++++++++++++++++++++
>> target-s390x/exec.h | 51 ++++++++++++++++++++
>> target-s390x/helper.c | 57 ++++++++++++++++++++++
>> target-s390x/op_helper.c | 74 ++++++++++++++++++++++++++++
>> target-s390x/translate.c | 57 ++++++++++++++++++++++
>> 6 files changed, 360 insertions(+), 0 deletions(-)
>> create mode 100644 target-s390x/cpu.h
>> create mode 100644 target-s390x/exec.h
>> create mode 100644 target-s390x/helper.c
>> create mode 100644 target-s390x/op_helper.c
>> create mode 100644 target-s390x/translate.c
>>
>> diff --git a/cpu-exec.c b/cpu-exec.c
>> index 2c0765c..af4595b 100644
>> --- a/cpu-exec.c
>> +++ b/cpu-exec.c
>> @@ -249,6 +249,7 @@ int cpu_exec(CPUState *env1)
>> #elif defined(TARGET_MIPS)
>> #elif defined(TARGET_SH4)
>> #elif defined(TARGET_CRIS)
>> +#elif defined(TARGET_S390X)
>> /* XXXXX */
>> #else
>> #error unsupported target CPU
>> @@ -673,6 +674,7 @@ int cpu_exec(CPUState *env1)
>> #elif defined(TARGET_SH4)
>> #elif defined(TARGET_ALPHA)
>> #elif defined(TARGET_CRIS)
>> +#elif defined(TARGET_S390X)
>> /* XXXXX */
>> #else
>> #error unsupported target CPU
>> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
>> new file mode 100644
>> index 0000000..f45b00c
>> --- /dev/null
>> +++ b/target-s390x/cpu.h
>> @@ -0,0 +1,119 @@
>> +/*
>> + * S/390 virtual CPU header
>> + *
>> + * Copyright (c) 2009 Ulrich Hecht
>> + *
>> + * This library is free software; you can redistribute it and/or
>> + * modify it under the terms of the GNU Lesser General Public
>> + * License as published by the Free Software Foundation; either
>> + * version 2 of the License, or (at your option) any later version.
>> + *
>> + * This library is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
>> + * Lesser General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU Lesser General Public
>> + * License along with this library; if not, write to the Free Software
>> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
>> + */
>> +#ifndef CPU_S390X_H
>> +#define CPU_S390X_H
>> +
>> +#define TARGET_LONG_BITS 64
>> +
>> +#define ELF_MACHINE EM_S390
>> +
>> +#define CPUState struct CPUS390XState
>> +
>> +#include "cpu-defs.h"
>> +
>> +#include "softfloat.h"
>> +
>> +#define NB_MMU_MODES 2 // guess
>> +#define MMU_USER_IDX 0 // guess
>> +
>> +typedef union FPReg {
>> + struct {
>> +#ifdef WORDS_BIGENDIAN
>> + float32 e;
>> + int32_t __pad;
>> +#else
>> + int32_t __pad;
>> + float32 e;
>> +#endif
>> + };
>> + float64 d;
>> + uint64_t i;
>> +} FPReg;
>> +
>> +typedef struct CPUS390XState {
>> + uint64_t regs[16]; /* GP registers */
>> +
>> + uint32_t aregs[16]; /* access registers */
>> +
>> + uint32_t fpc; /* floating-point control register */
>> + FPReg fregs[16]; /* FP registers */
>> + float_status fpu_status; /* passed to softfloat lib */
>> +
>> + struct {
>> + uint64_t mask;
>> + uint64_t addr;
>> + } psw;
>> +
>> + int cc; /* condition code (0-3) */
>> +
>> + uint64_t __excp_addr;
>> +
>> + CPU_COMMON
>> +} CPUS390XState;
>> +
>> +#if defined(CONFIG_USER_ONLY)
>> +static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
>> +{
>> + if (newsp)
>> + env->regs[15] = newsp;
>
> Coding style.
>
>> + env->regs[0] = 0;
>> +}
>> +#endif
>> +
>> +CPUS390XState *cpu_s390x_init(const char *cpu_model);
>> +int cpu_s390x_exec(CPUS390XState *s);
>> +void cpu_s390x_close(CPUS390XState *s);
>> +
>> +/* you can call this signal handler from your SIGBUS and SIGSEGV
>> + signal handlers to inform the virtual CPU of exceptions. non zero
>> + is returned if the signal was handled by the virtual CPU. */
>> +int cpu_s390x_signal_handler(int host_signum, void *pinfo,
>> + void *puc);
>> +int cpu_s390x_handle_mmu_fault (CPUS390XState *env, target_ulong address, int rw,
>> + int mmu_idx, int is_softmuu);
>> +#define cpu_handle_mmu_fault cpu_s390x_handle_mmu_fault
>> +
>> +#define TARGET_PAGE_BITS 12
>> +
>> +#define cpu_init cpu_s390x_init
>> +#define cpu_exec cpu_s390x_exec
>> +#define cpu_gen_code cpu_s390x_gen_code
>> +
>> +#include "cpu-all.h"
>> +#include "exec-all.h"
>> +
>> +#define EXCP_OPEX 1 /* operation exception (sigill) */
>> +#define EXCP_SVC 2 /* supervisor call (syscall) */
>> +#define EXCP_ADDR 5 /* addressing exception */
>> +#define EXCP_EXECUTE_SVC 0xff00000 /* supervisor call via execute insn */
>> +
>> +static inline void cpu_pc_from_tb(CPUState *env, TranslationBlock* tb)
>> +{
>> + env->psw.addr = tb->pc;
>> +}
>> +
>> +static inline void cpu_get_tb_cpu_state(CPUState* env, target_ulong *pc,
>> + target_ulong *cs_base, int *flags)
>> +{
>> + *pc = env->psw.addr;
>> + *cs_base = 0;
>> + *flags = env->psw.mask; // guess
>
> I don't know what psw.mask represent, but it may be wrong. It should be
> a way to identify which TB can be reused, that is they have been
> generated in the same CPU mode.
psw.mask is rougly the same as RFLAGS, cr0 and cr4 on x86_64 combined. So IMHO it looked like a pretty good identifier for TB uniqueness.
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-11-30 22:30 ` Alexander Graf
@ 2009-12-01 9:46 ` Carsten Otte
2009-12-01 10:11 ` Alexander Graf
2009-12-01 18:45 ` Blue Swirl
2009-12-02 8:17 ` Aurelien Jarno
1 sibling, 2 replies; 49+ messages in thread
From: Carsten Otte @ 2009-12-01 9:46 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel, Aurelien Jarno
Alexander Graf wrote:
>> I don't know what psw.mask represent, but it may be wrong. It should be
>> a way to identify which TB can be reused, that is they have been
>> generated in the same CPU mode.
>
> psw.mask is rougly the same as RFLAGS, cr0 and cr4 on x86_64 combined. So IMHO it looked like a pretty good identifier for TB uniqueness.
I am not familar with qemu at all here, therefore the following
explanation may not fit here. I assume the translation block refers to
guest virtual to guest physical memory translations. In that case this
is not the right indicator on it's own. The right indicator which
translation the cpu would do would be pretty complex:
Our cpu keeps multiple seperate address spaces open at the same time
(similar to x86 with a bunch of cr0s), defined by address space control
elements in various control registers. Linux uses primary, secondary and
home space to address user space and kernel space. The third one is user
space once again for exec-type access (to implement stack execute
protection). PSW.mask selects which one is to be used for address
translation by _default_. Even worse, the cpu may load instructions and
data from different adddress spaces (secondary space mdoe). Yet more
worse some instructions use "access register mode" where a general
purpose register points to yet another address space. A detailed
documentation can be found here:
http://publibfp.boulder.ibm.com/cgi-bin/bookmgr/BOOKS/dz9zr002/3.0?DT=20030424140649
That said, I think it's best to keep out softmmu for now. It's not
needed for kvm operation and very complex to do right.
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-12-01 9:46 ` Carsten Otte
@ 2009-12-01 10:11 ` Alexander Graf
2009-12-01 18:45 ` Blue Swirl
1 sibling, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-12-01 10:11 UTC (permalink / raw)
To: Carsten Otte; +Cc: Carsten Otte, qemu-devel, Aurelien Jarno
On 01.12.2009, at 10:46, Carsten Otte wrote:
> Alexander Graf wrote:
>>> I don't know what psw.mask represent, but it may be wrong. It should be
>>> a way to identify which TB can be reused, that is they have been
>>> generated in the same CPU mode.
>> psw.mask is rougly the same as RFLAGS, cr0 and cr4 on x86_64 combined. So IMHO it looked like a pretty good identifier for TB uniqueness.
> I am not familar with qemu at all here, therefore the following explanation may not fit here. I assume the translation block refers to guest virtual to guest physical memory translations. In that case this is not the right indicator on it's own. The right indicator which translation the cpu would do would be pretty complex:
> Our cpu keeps multiple seperate address spaces open at the same time (similar to x86 with a bunch of cr0s), defined by address space control elements in various control registers. Linux uses primary, secondary and home space to address user space and kernel space. The third one is user space once again for exec-type access (to implement stack execute protection). PSW.mask selects which one is to be used for address translation by _default_. Even worse, the cpu may load instructions and data from different adddress spaces (secondary space mdoe). Yet more worse some instructions use "access register mode" where a general purpose register points to yet another address space. A detailed documentation can be found here: http://publibfp.boulder.ibm.com/cgi-bin/bookmgr/BOOKS/dz9zr002/3.0?DT=20030424140649
>
> That said, I think it's best to keep out softmmu for now. It's not needed for kvm operation and very complex to do right.
Ah ok, so we're missing the asce register information for softmmu uniqueness. Well - I guess I should put a comment there that tells anyone who feels like implementing a softmmu target that this is a pitfall.
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-12-01 9:46 ` Carsten Otte
2009-12-01 10:11 ` Alexander Graf
@ 2009-12-01 18:45 ` Blue Swirl
2009-12-02 7:55 ` Carsten Otte
2009-12-02 14:41 ` Paul Brook
1 sibling, 2 replies; 49+ messages in thread
From: Blue Swirl @ 2009-12-01 18:45 UTC (permalink / raw)
To: Carsten Otte; +Cc: Carsten Otte, Alexander Graf, Aurelien Jarno, qemu-devel
2009/12/1 Carsten Otte <carsteno@de.ibm.com>:
> Alexander Graf wrote:
>>>
>>> I don't know what psw.mask represent, but it may be wrong. It should be
>>> a way to identify which TB can be reused, that is they have been
>>> generated in the same CPU mode.
>>
>> psw.mask is rougly the same as RFLAGS, cr0 and cr4 on x86_64 combined. So
>> IMHO it looked like a pretty good identifier for TB uniqueness.
>
> I am not familar with qemu at all here, therefore the following explanation
> may not fit here. I assume the translation block refers to guest virtual to
> guest physical memory translations. In that case this is not the right
> indicator on it's own. The right indicator which translation the cpu would
> do would be pretty complex:
The translation block (TB) refers to a block of host instructions,
translated from some block of target instructions under some
assumptions. The assumptions used when translating (for example, user
vs supervisor mode in the CPU state) are recorded to TB flags. If the
CPU state changes, the old TBs with different flags will not be used
and new TBs will be generated.
> Our cpu keeps multiple seperate address spaces open at the same time
> (similar to x86 with a bunch of cr0s), defined by address space control
> elements in various control registers. Linux uses primary, secondary and
> home space to address user space and kernel space. The third one is user
> space once again for exec-type access (to implement stack execute
> protection). PSW.mask selects which one is to be used for address
> translation by _default_. Even worse, the cpu may load instructions and data
> from different adddress spaces (secondary space mdoe). Yet more worse some
> instructions use "access register mode" where a general purpose register
> points to yet another address space. A detailed documentation can be found
> here:
> http://publibfp.boulder.ibm.com/cgi-bin/bookmgr/BOOKS/dz9zr002/3.0?DT=20030424140649
Actually Sparc64 address spaces and ASIs are very similar. There are
nucleus, primary and secondary address spaces (not fully implemented
yet in QEMU). Instructions can encode the ASI or %asi register can be
used. Some ASIs are restricted for supervisor or hypervisor modes.
Sparc32 ASIs are simpler (physical address space extension to 36 bits,
basically) and for supervisor only.
For S/390, I think the TB flags do not need to contain the address
space control registers if the generated instructions fetch the state
from CPU state and do not rely on translation time information. If the
address spaces do not change very often, it may alternatively be
possible to rely on the CPU state during translation, but then it must
be ensured that all generated TBs are always flushed when the
registers change.
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-12-01 18:45 ` Blue Swirl
@ 2009-12-02 7:55 ` Carsten Otte
2009-12-02 14:41 ` Paul Brook
1 sibling, 0 replies; 49+ messages in thread
From: Carsten Otte @ 2009-12-02 7:55 UTC (permalink / raw)
To: Blue Swirl; +Cc: Carsten Otte, Alexander Graf, Aurelien Jarno, qemu-devel
Blue Swirl wrote:
> The translation block (TB) refers to a block of host instructions,
> translated from some block of target instructions under some
> assumptions. The assumptions used when translating (for example, user
> vs supervisor mode in the CPU state) are recorded to TB flags. If the
> CPU state changes, the old TBs with different flags will not be used
> and new TBs will be generated.
Thanks for the explanation, Alex did explain it too on irc.
> Actually Sparc64 address spaces and ASIs are very similar. There are
> nucleus, primary and secondary address spaces (not fully implemented
> yet in QEMU). Instructions can encode the ASI or %asi register can be
> used. Some ASIs are restricted for supervisor or hypervisor modes.
> Sparc32 ASIs are simpler (physical address space extension to 36 bits,
> basically) and for supervisor only.
Oh, I though we were the only odd bird on this one.
> For S/390, I think the TB flags do not need to contain the address
> space control registers if the generated instructions fetch the state
> from CPU state and do not rely on translation time information. If the
> address spaces do not change very often, it may alternatively be
> possible to rely on the CPU state during translation, but then it must
> be ensured that all generated TBs are always flushed when the
> registers change.
Yes, that would work as far as I can tell. The registers change whenever
$guest decides to schedule a different task, but not on every syscall.
cheers,
Carsten
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-12-01 18:45 ` Blue Swirl
2009-12-02 7:55 ` Carsten Otte
@ 2009-12-02 14:41 ` Paul Brook
1 sibling, 0 replies; 49+ messages in thread
From: Paul Brook @ 2009-12-02 14:41 UTC (permalink / raw)
To: qemu-devel
Cc: Blue Swirl, Carsten Otte, Alexander Graf, Aurelien Jarno,
Carsten Otte
> > Our cpu keeps multiple seperate address spaces open at the same time
> > (similar to x86 with a bunch of cr0s), defined by address space control
> > elements in various control registers. Linux uses primary, secondary and
> > home space to address user space and kernel space. The third one is user
> > space once again for exec-type access (to implement stack execute
> > protection). PSW.mask selects which one is to be used for address
> > translation by _default_. Even worse, the cpu may load instructions and
> > data from different adddress spaces (secondary space mdoe). Yet more
> > worse some instructions use "access register mode" where a general
> > purpose register points to yet another address space. A detailed
> > documentation can be found here:
> > http://publibfp.boulder.ibm.com/cgi-bin/bookmgr/BOOKS/dz9zr002/3.0?DT=200
> >30424140649
>
> Actually Sparc64 address spaces and ASIs are very similar. There are
> nucleus, primary and secondary address spaces (not fully implemented
> yet in QEMU). Instructions can encode the ASI or %asi register can be
> used. Some ASIs are restricted for supervisor or hypervisor modes.
> Sparc32 ASIs are simpler (physical address space extension to 36 bits,
> basically) and for supervisor only.
>
> For S/390, I think the TB flags do not need to contain the address
> space control registers if the generated instructions fetch the state
> from CPU state and do not rely on translation time information. If the
> address spaces do not change very often, it may alternatively be
> possible to rely on the CPU state during translation, but then it must
> be ensured that all generated TBs are always flushed when the
> registers change.
It sounds like there's some confusion between virtual address translation and
state that effects instruction semantics. The TB flags should include the
latter. The former isn't particularly well supported in qemu. If you have more
than a couple of different address spaces (i.e. kernel and userspace) then you
basically have to flush the TLB every time the current address space changes.
If the address space an be selected per-instruction then you're pretty much
screwed.
Paul
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-11-30 22:30 ` Alexander Graf
2009-12-01 9:46 ` Carsten Otte
@ 2009-12-02 8:17 ` Aurelien Jarno
2009-12-02 8:30 ` Alexander Graf
1 sibling, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-12-02 8:17 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Mon, Nov 30, 2009 at 11:30:04PM +0100, Alexander Graf wrote:
>
> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>
> > On Thu, Nov 26, 2009 at 02:23:10PM +0100, Alexander Graf wrote:
> >> Because Qemu currently requires a TCG target to exist and there are quite some
> >> useful helpers here to lay the groundwork for out KVM target, let's create a
> >> stub TCG emulation target for S390X CPUs.
> >>
> >> This is required to make tcg happy. The emulation target itself won't work
> >> though.
> >
> > Please find the comments below.
> >
> >> Signed-off-by: Alexander Graf <agraf@suse.de>
> >> ---
> >> cpu-exec.c | 2 +
> >> target-s390x/cpu.h | 119 ++++++++++++++++++++++++++++++++++++++++++++++
> >> target-s390x/exec.h | 51 ++++++++++++++++++++
> >> target-s390x/helper.c | 57 ++++++++++++++++++++++
> >> target-s390x/op_helper.c | 74 ++++++++++++++++++++++++++++
> >> target-s390x/translate.c | 57 ++++++++++++++++++++++
> >> 6 files changed, 360 insertions(+), 0 deletions(-)
> >> create mode 100644 target-s390x/cpu.h
> >> create mode 100644 target-s390x/exec.h
> >> create mode 100644 target-s390x/helper.c
> >> create mode 100644 target-s390x/op_helper.c
> >> create mode 100644 target-s390x/translate.c
> >>
> >> diff --git a/cpu-exec.c b/cpu-exec.c
> >> index 2c0765c..af4595b 100644
> >> --- a/cpu-exec.c
> >> +++ b/cpu-exec.c
> >> @@ -249,6 +249,7 @@ int cpu_exec(CPUState *env1)
> >> #elif defined(TARGET_MIPS)
> >> #elif defined(TARGET_SH4)
> >> #elif defined(TARGET_CRIS)
> >> +#elif defined(TARGET_S390X)
> >> /* XXXXX */
> >> #else
> >> #error unsupported target CPU
> >> @@ -673,6 +674,7 @@ int cpu_exec(CPUState *env1)
> >> #elif defined(TARGET_SH4)
> >> #elif defined(TARGET_ALPHA)
> >> #elif defined(TARGET_CRIS)
> >> +#elif defined(TARGET_S390X)
> >> /* XXXXX */
> >> #else
> >> #error unsupported target CPU
> >> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
> >> new file mode 100644
> >> index 0000000..f45b00c
> >> --- /dev/null
> >> +++ b/target-s390x/cpu.h
> >> @@ -0,0 +1,119 @@
> >> +/*
> >> + * S/390 virtual CPU header
> >> + *
> >> + * Copyright (c) 2009 Ulrich Hecht
> >> + *
> >> + * This library is free software; you can redistribute it and/or
> >> + * modify it under the terms of the GNU Lesser General Public
> >> + * License as published by the Free Software Foundation; either
> >> + * version 2 of the License, or (at your option) any later version.
> >> + *
> >> + * This library is distributed in the hope that it will be useful,
> >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> >> + * Lesser General Public License for more details.
> >> + *
> >> + * You should have received a copy of the GNU Lesser General Public
> >> + * License along with this library; if not, write to the Free Software
> >> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
> >> + */
> >> +#ifndef CPU_S390X_H
> >> +#define CPU_S390X_H
> >> +
> >> +#define TARGET_LONG_BITS 64
> >> +
> >> +#define ELF_MACHINE EM_S390
> >> +
> >> +#define CPUState struct CPUS390XState
> >> +
> >> +#include "cpu-defs.h"
> >> +
> >> +#include "softfloat.h"
> >> +
> >> +#define NB_MMU_MODES 2 // guess
> >> +#define MMU_USER_IDX 0 // guess
> >> +
> >> +typedef union FPReg {
> >> + struct {
> >> +#ifdef WORDS_BIGENDIAN
> >> + float32 e;
> >> + int32_t __pad;
> >> +#else
> >> + int32_t __pad;
> >> + float32 e;
> >> +#endif
> >> + };
> >> + float64 d;
> >> + uint64_t i;
> >> +} FPReg;
> >> +
> >> +typedef struct CPUS390XState {
> >> + uint64_t regs[16]; /* GP registers */
> >> +
> >> + uint32_t aregs[16]; /* access registers */
> >> +
> >> + uint32_t fpc; /* floating-point control register */
> >> + FPReg fregs[16]; /* FP registers */
> >> + float_status fpu_status; /* passed to softfloat lib */
> >> +
> >> + struct {
> >> + uint64_t mask;
> >> + uint64_t addr;
> >> + } psw;
> >> +
> >> + int cc; /* condition code (0-3) */
> >> +
> >> + uint64_t __excp_addr;
> >> +
> >> + CPU_COMMON
> >> +} CPUS390XState;
> >> +
> >> +#if defined(CONFIG_USER_ONLY)
> >> +static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
> >> +{
> >> + if (newsp)
> >> + env->regs[15] = newsp;
> >
> > Coding style.
> >
> >> + env->regs[0] = 0;
> >> +}
> >> +#endif
> >> +
> >> +CPUS390XState *cpu_s390x_init(const char *cpu_model);
> >> +int cpu_s390x_exec(CPUS390XState *s);
> >> +void cpu_s390x_close(CPUS390XState *s);
> >> +
> >> +/* you can call this signal handler from your SIGBUS and SIGSEGV
> >> + signal handlers to inform the virtual CPU of exceptions. non zero
> >> + is returned if the signal was handled by the virtual CPU. */
> >> +int cpu_s390x_signal_handler(int host_signum, void *pinfo,
> >> + void *puc);
> >> +int cpu_s390x_handle_mmu_fault (CPUS390XState *env, target_ulong address, int rw,
> >> + int mmu_idx, int is_softmuu);
> >> +#define cpu_handle_mmu_fault cpu_s390x_handle_mmu_fault
> >> +
> >> +#define TARGET_PAGE_BITS 12
> >> +
> >> +#define cpu_init cpu_s390x_init
> >> +#define cpu_exec cpu_s390x_exec
> >> +#define cpu_gen_code cpu_s390x_gen_code
> >> +
> >> +#include "cpu-all.h"
> >> +#include "exec-all.h"
> >> +
> >> +#define EXCP_OPEX 1 /* operation exception (sigill) */
> >> +#define EXCP_SVC 2 /* supervisor call (syscall) */
> >> +#define EXCP_ADDR 5 /* addressing exception */
> >> +#define EXCP_EXECUTE_SVC 0xff00000 /* supervisor call via execute insn */
> >> +
> >> +static inline void cpu_pc_from_tb(CPUState *env, TranslationBlock* tb)
> >> +{
> >> + env->psw.addr = tb->pc;
> >> +}
> >> +
> >> +static inline void cpu_get_tb_cpu_state(CPUState* env, target_ulong *pc,
> >> + target_ulong *cs_base, int *flags)
> >> +{
> >> + *pc = env->psw.addr;
> >> + *cs_base = 0;
> >> + *flags = env->psw.mask; // guess
> >
> > I don't know what psw.mask represent, but it may be wrong. It should be
> > a way to identify which TB can be reused, that is they have been
> > generated in the same CPU mode.
>
> psw.mask is rougly the same as RFLAGS, cr0 and cr4 on x86_64 combined. So IMHO it looked like a pretty good identifier for TB uniqueness.
>
Then it's most probably correct, except some bits may have to be
ignored. Can you update the comment explaining that?
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation
2009-12-02 8:17 ` Aurelien Jarno
@ 2009-12-02 8:30 ` Alexander Graf
0 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-12-02 8:30 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 02.12.2009, at 09:17, Aurelien Jarno wrote:
> On Mon, Nov 30, 2009 at 11:30:04PM +0100, Alexander Graf wrote:
>>
>> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>>
>>> On Thu, Nov 26, 2009 at 02:23:10PM +0100, Alexander Graf wrote:
>>>> Because Qemu currently requires a TCG target to exist and there are quite some
>>>> useful helpers here to lay the groundwork for out KVM target, let's create a
>>>> stub TCG emulation target for S390X CPUs.
>>>>
>>>> This is required to make tcg happy. The emulation target itself won't work
>>>> though.
>>>
>>> Please find the comments below.
>>>
>>>> Signed-off-by: Alexander Graf <agraf@suse.de>
>>>> ---
>>>> cpu-exec.c | 2 +
>>>> target-s390x/cpu.h | 119 ++++++++++++++++++++++++++++++++++++++++++++++
>>>> target-s390x/exec.h | 51 ++++++++++++++++++++
>>>> target-s390x/helper.c | 57 ++++++++++++++++++++++
>>>> target-s390x/op_helper.c | 74 ++++++++++++++++++++++++++++
>>>> target-s390x/translate.c | 57 ++++++++++++++++++++++
>>>> 6 files changed, 360 insertions(+), 0 deletions(-)
>>>> create mode 100644 target-s390x/cpu.h
>>>> create mode 100644 target-s390x/exec.h
>>>> create mode 100644 target-s390x/helper.c
>>>> create mode 100644 target-s390x/op_helper.c
>>>> create mode 100644 target-s390x/translate.c
>>>>
>>>> diff --git a/cpu-exec.c b/cpu-exec.c
>>>> index 2c0765c..af4595b 100644
>>>> --- a/cpu-exec.c
>>>> +++ b/cpu-exec.c
>>>> @@ -249,6 +249,7 @@ int cpu_exec(CPUState *env1)
>>>> #elif defined(TARGET_MIPS)
>>>> #elif defined(TARGET_SH4)
>>>> #elif defined(TARGET_CRIS)
>>>> +#elif defined(TARGET_S390X)
>>>> /* XXXXX */
>>>> #else
>>>> #error unsupported target CPU
>>>> @@ -673,6 +674,7 @@ int cpu_exec(CPUState *env1)
>>>> #elif defined(TARGET_SH4)
>>>> #elif defined(TARGET_ALPHA)
>>>> #elif defined(TARGET_CRIS)
>>>> +#elif defined(TARGET_S390X)
>>>> /* XXXXX */
>>>> #else
>>>> #error unsupported target CPU
>>>> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
>>>> new file mode 100644
>>>> index 0000000..f45b00c
>>>> --- /dev/null
>>>> +++ b/target-s390x/cpu.h
>>>> @@ -0,0 +1,119 @@
>>>> +/*
>>>> + * S/390 virtual CPU header
>>>> + *
>>>> + * Copyright (c) 2009 Ulrich Hecht
>>>> + *
>>>> + * This library is free software; you can redistribute it and/or
>>>> + * modify it under the terms of the GNU Lesser General Public
>>>> + * License as published by the Free Software Foundation; either
>>>> + * version 2 of the License, or (at your option) any later version.
>>>> + *
>>>> + * This library is distributed in the hope that it will be useful,
>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
>>>> + * Lesser General Public License for more details.
>>>> + *
>>>> + * You should have received a copy of the GNU Lesser General Public
>>>> + * License along with this library; if not, write to the Free Software
>>>> + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301 USA
>>>> + */
>>>> +#ifndef CPU_S390X_H
>>>> +#define CPU_S390X_H
>>>> +
>>>> +#define TARGET_LONG_BITS 64
>>>> +
>>>> +#define ELF_MACHINE EM_S390
>>>> +
>>>> +#define CPUState struct CPUS390XState
>>>> +
>>>> +#include "cpu-defs.h"
>>>> +
>>>> +#include "softfloat.h"
>>>> +
>>>> +#define NB_MMU_MODES 2 // guess
>>>> +#define MMU_USER_IDX 0 // guess
>>>> +
>>>> +typedef union FPReg {
>>>> + struct {
>>>> +#ifdef WORDS_BIGENDIAN
>>>> + float32 e;
>>>> + int32_t __pad;
>>>> +#else
>>>> + int32_t __pad;
>>>> + float32 e;
>>>> +#endif
>>>> + };
>>>> + float64 d;
>>>> + uint64_t i;
>>>> +} FPReg;
>>>> +
>>>> +typedef struct CPUS390XState {
>>>> + uint64_t regs[16]; /* GP registers */
>>>> +
>>>> + uint32_t aregs[16]; /* access registers */
>>>> +
>>>> + uint32_t fpc; /* floating-point control register */
>>>> + FPReg fregs[16]; /* FP registers */
>>>> + float_status fpu_status; /* passed to softfloat lib */
>>>> +
>>>> + struct {
>>>> + uint64_t mask;
>>>> + uint64_t addr;
>>>> + } psw;
>>>> +
>>>> + int cc; /* condition code (0-3) */
>>>> +
>>>> + uint64_t __excp_addr;
>>>> +
>>>> + CPU_COMMON
>>>> +} CPUS390XState;
>>>> +
>>>> +#if defined(CONFIG_USER_ONLY)
>>>> +static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
>>>> +{
>>>> + if (newsp)
>>>> + env->regs[15] = newsp;
>>>
>>> Coding style.
>>>
>>>> + env->regs[0] = 0;
>>>> +}
>>>> +#endif
>>>> +
>>>> +CPUS390XState *cpu_s390x_init(const char *cpu_model);
>>>> +int cpu_s390x_exec(CPUS390XState *s);
>>>> +void cpu_s390x_close(CPUS390XState *s);
>>>> +
>>>> +/* you can call this signal handler from your SIGBUS and SIGSEGV
>>>> + signal handlers to inform the virtual CPU of exceptions. non zero
>>>> + is returned if the signal was handled by the virtual CPU. */
>>>> +int cpu_s390x_signal_handler(int host_signum, void *pinfo,
>>>> + void *puc);
>>>> +int cpu_s390x_handle_mmu_fault (CPUS390XState *env, target_ulong address, int rw,
>>>> + int mmu_idx, int is_softmuu);
>>>> +#define cpu_handle_mmu_fault cpu_s390x_handle_mmu_fault
>>>> +
>>>> +#define TARGET_PAGE_BITS 12
>>>> +
>>>> +#define cpu_init cpu_s390x_init
>>>> +#define cpu_exec cpu_s390x_exec
>>>> +#define cpu_gen_code cpu_s390x_gen_code
>>>> +
>>>> +#include "cpu-all.h"
>>>> +#include "exec-all.h"
>>>> +
>>>> +#define EXCP_OPEX 1 /* operation exception (sigill) */
>>>> +#define EXCP_SVC 2 /* supervisor call (syscall) */
>>>> +#define EXCP_ADDR 5 /* addressing exception */
>>>> +#define EXCP_EXECUTE_SVC 0xff00000 /* supervisor call via execute insn */
>>>> +
>>>> +static inline void cpu_pc_from_tb(CPUState *env, TranslationBlock* tb)
>>>> +{
>>>> + env->psw.addr = tb->pc;
>>>> +}
>>>> +
>>>> +static inline void cpu_get_tb_cpu_state(CPUState* env, target_ulong *pc,
>>>> + target_ulong *cs_base, int *flags)
>>>> +{
>>>> + *pc = env->psw.addr;
>>>> + *cs_base = 0;
>>>> + *flags = env->psw.mask; // guess
>>>
>>> I don't know what psw.mask represent, but it may be wrong. It should be
>>> a way to identify which TB can be reused, that is they have been
>>> generated in the same CPU mode.
>>
>> psw.mask is rougly the same as RFLAGS, cr0 and cr4 on x86_64 combined. So IMHO it looked like a pretty good identifier for TB uniqueness.
>>
>
> Then it's most probably correct, except some bits may have to be
> ignored. Can you update the comment explaining that?
Yep.
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread
* [Qemu-devel] [PATCH 02/11] S/390 host/target build system support
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
2009-11-26 13:23 ` [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-26 13:23 ` [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation Alexander Graf
` (9 subsequent siblings)
11 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
This patch makes configure aware of S390 hosts and guests. When not explicitly
defined using --target-list= no S390 targets will be built though.
Signed-off-by: Alexander Graf <agraf@suse.de>
---
configure | 20 ++++++++++++++------
1 files changed, 14 insertions(+), 6 deletions(-)
diff --git a/configure b/configure
index 1223fc8..b616e1a 100755
--- a/configure
+++ b/configure
@@ -158,9 +158,12 @@ case "$cpu" in
parisc|parisc64)
cpu="hppa"
;;
- s390*)
+ s390)
cpu="s390"
;;
+ s390x)
+ cpu="s390x"
+ ;;
sparc|sun4[cdmuv])
cpu="sparc"
;;
@@ -862,7 +865,7 @@ fi
# host long bits test
hostlongbits="32"
case "$cpu" in
- x86_64|alpha|ia64|sparc64|ppc64)
+ x86_64|alpha|ia64|sparc64|ppc64|s390x)
hostlongbits=64
;;
esac
@@ -1874,7 +1877,7 @@ echo >> $config_host_mak
echo "CONFIG_QEMU_SHAREDIR=\"$prefix$datasuffix\"" >> $config_host_mak
case "$cpu" in
- i386|x86_64|alpha|cris|hppa|ia64|m68k|microblaze|mips|mips64|ppc|ppc64|s390|sparc|sparc64)
+ i386|x86_64|alpha|cris|hppa|ia64|m68k|microblaze|mips|mips64|ppc|ppc64|s390|s390x|sparc|sparc64)
ARCH=$cpu
;;
armv4b|armv4l)
@@ -2158,7 +2161,7 @@ target_arch2=`echo $target | cut -d '-' -f 1`
target_bigendian="no"
case "$target_arch2" in
- armeb|m68k|microblaze|mips|mipsn32|mips64|ppc|ppcemb|ppc64|ppc64abi32|sh4eb|sparc|sparc64|sparc32plus)
+ armeb|m68k|microblaze|mips|mipsn32|mips64|ppc|ppcemb|ppc64|ppc64abi32|s390x|sh4eb|sparc|sparc64|sparc32plus)
target_bigendian=yes
;;
esac
@@ -2326,6 +2329,9 @@ case "$target_arch2" in
echo "TARGET_ABI32=y" >> $config_target_mak
target_phys_bits=64
;;
+ s390x)
+ target_phys_bits=64
+ ;;
*)
echo "Unsupported target CPU"
exit 1
@@ -2394,7 +2400,7 @@ if test ! -z "$gdb_xml_files" ; then
fi
case "$target_arch2" in
- arm|armeb|m68k|microblaze|mips|mipsel|mipsn32|mipsn32el|mips64|mips64el|ppc|ppc64|ppc64abi32|ppcemb|sparc|sparc64|sparc32plus)
+ arm|armeb|m68k|microblaze|mips|mipsel|mipsn32|mipsn32el|mips64|mips64el|ppc|ppc64|ppc64abi32|ppcemb|s390x|sparc|sparc64|sparc32plus)
echo "CONFIG_SOFTFLOAT=y" >> $config_target_mak
;;
*)
@@ -2427,6 +2433,8 @@ ldflags=""
if test "$ARCH" = "sparc64" ; then
cflags="-I\$(SRC_PATH)/tcg/sparc $cflags"
+elif test "$ARCH" = "s390x" ; then
+ cflags="-I\$(SRC_PATH)/tcg/s390 $cflags"
else
cflags="-I\$(SRC_PATH)/tcg/\$(ARCH) $cflags"
fi
@@ -2462,7 +2470,7 @@ for i in $ARCH $TARGET_BASE_ARCH ; do
ppc*)
echo "CONFIG_PPC_DIS=y" >> $config_target_mak
;;
- s390)
+ s390*)
echo "CONFIG_S390_DIS=y" >> $config_target_mak
;;
sh4)
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
2009-11-26 13:23 ` [Qemu-devel] [PATCH 01/11] S/390 CPU fake emulation Alexander Graf
2009-11-26 13:23 ` [Qemu-devel] [PATCH 02/11] S/390 host/target build system support Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-30 18:18 ` Aurelien Jarno
2009-11-26 13:23 ` [Qemu-devel] [PATCH 04/11] Add KVM support for S390x Alexander Graf
` (8 subsequent siblings)
11 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
Qemu won't let us run a KVM target without having host TCG support. Well, for
now we don't have any so let's implement a fake target that only stubs out
everything.
I tried to keep the patch as close to Uli's source as possible, so whenever
he feels like it he can easily diff his version against this one.
Signed-off-by: Alexander Graf <agraf@suse.de>
---
dyngen-exec.h | 2 +-
target-s390x/helper.c | 5 ++
tcg/s390/tcg-target.c | 103 +++++++++++++++++++++++++++++++++++++++++++++++++
tcg/s390/tcg-target.h | 48 +++++++++++++++++++++++
4 files changed, 157 insertions(+), 1 deletions(-)
create mode 100644 tcg/s390/tcg-target.c
create mode 100644 tcg/s390/tcg-target.h
diff --git a/dyngen-exec.h b/dyngen-exec.h
index 86e61c3..0353f36 100644
--- a/dyngen-exec.h
+++ b/dyngen-exec.h
@@ -117,7 +117,7 @@ extern int printf(const char *, ...);
/* The return address may point to the start of the next instruction.
Subtracting one gets us the call instruction itself. */
-#if defined(__s390__)
+#if defined(__s390__) && !defined(__s390x__)
# define GETPC() ((void*)(((unsigned long)__builtin_return_address(0) & 0x7fffffffUL) - 1))
#elif defined(__arm__)
/* Thumb return addresses have the low bit set, so we need to subtract two.
diff --git a/target-s390x/helper.c b/target-s390x/helper.c
index 4e23b4a..0e222e3 100644
--- a/target-s390x/helper.c
+++ b/target-s390x/helper.c
@@ -44,6 +44,11 @@ CPUS390XState *cpu_s390x_init(const char *cpu_model)
return env;
}
+target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
+{
+ return addr;
+}
+
void cpu_reset(CPUS390XState *env)
{
if (qemu_loglevel_mask(CPU_LOG_RESET)) {
diff --git a/tcg/s390/tcg-target.c b/tcg/s390/tcg-target.c
new file mode 100644
index 0000000..53bbf69
--- /dev/null
+++ b/tcg/s390/tcg-target.c
@@ -0,0 +1,103 @@
+/*
+ * Tiny Code Generator for QEMU
+ *
+ * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+static const int tcg_target_reg_alloc_order[] = {
+};
+
+static const int tcg_target_call_iarg_regs[] = {
+};
+
+static const int tcg_target_call_oarg_regs[] = {
+};
+
+static void patch_reloc(uint8_t *code_ptr, int type,
+ tcg_target_long value, tcg_target_long addend)
+{
+ tcg_abort();
+}
+
+static inline int tcg_target_get_call_iarg_regs_count(int flags)
+{
+ tcg_abort();
+ return 0;
+}
+
+/* parse target specific constraints */
+static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
+{
+ tcg_abort();
+ return 0;
+}
+
+/* Test if a constant matches the constraint. */
+static inline int tcg_target_const_match(tcg_target_long val,
+ const TCGArgConstraint *arg_ct)
+{
+ tcg_abort();
+ return 0;
+}
+
+/* load a register with an immediate value */
+static inline void tcg_out_movi(TCGContext *s, TCGType type,
+ int ret, tcg_target_long arg)
+{
+ tcg_abort();
+}
+
+/* load data without address translation or endianness conversion */
+static inline void tcg_out_ld(TCGContext *s, TCGType type, int arg,
+ int arg1, tcg_target_long arg2)
+{
+ tcg_abort();
+}
+
+static inline void tcg_out_st(TCGContext *s, TCGType type, int arg,
+ int arg1, tcg_target_long arg2)
+{
+ tcg_abort();
+}
+
+static inline void tcg_out_op(TCGContext *s, int opc,
+ const TCGArg *args, const int *const_args)
+{
+ tcg_abort();
+}
+
+void tcg_target_init(TCGContext *s)
+{
+}
+
+void tcg_target_qemu_prologue(TCGContext *s)
+{
+}
+
+static inline void tcg_out_mov(TCGContext *s, int ret, int arg)
+{
+ tcg_abort();
+}
+
+static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val)
+{
+ tcg_abort();
+}
diff --git a/tcg/s390/tcg-target.h b/tcg/s390/tcg-target.h
new file mode 100644
index 0000000..2e66553
--- /dev/null
+++ b/tcg/s390/tcg-target.h
@@ -0,0 +1,48 @@
+/*
+ * Tiny Code Generator for QEMU
+ *
+ * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+#define TCG_TARGET_S390 1
+
+#define TCG_TARGET_REG_BITS 64
+#define TCG_TARGET_WORDS_BIGENDIAN
+
+#define TCG_TARGET_NB_REGS 0
+
+enum {
+ TCG_AREG0 = 0,
+};
+
+/* used for function call generation */
+#define TCG_REG_CALL_STACK 0
+#define TCG_TARGET_STACK_ALIGN 8
+#define TCG_TARGET_CALL_STACK_OFFSET 0
+
+static inline void flush_icache_range(unsigned long start, unsigned long stop)
+{
+#if QEMU_GNUC_PREREQ(4, 1)
+ void __clear_cache(char *beg, char *end);
+ __clear_cache((char *) start, (char *) stop);
+#else
+#error not implemented
+#endif
+}
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation
2009-11-26 13:23 ` [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation Alexander Graf
@ 2009-11-30 18:18 ` Aurelien Jarno
2009-11-30 22:27 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-11-30 18:18 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Thu, Nov 26, 2009 at 02:23:12PM +0100, Alexander Graf wrote:
> Qemu won't let us run a KVM target without having host TCG support. Well, for
> now we don't have any so let's implement a fake target that only stubs out
> everything.
>
> I tried to keep the patch as close to Uli's source as possible, so whenever
> he feels like it he can easily diff his version against this one.
Please find the comments below.
> Signed-off-by: Alexander Graf <agraf@suse.de>
> ---
> dyngen-exec.h | 2 +-
> target-s390x/helper.c | 5 ++
> tcg/s390/tcg-target.c | 103 +++++++++++++++++++++++++++++++++++++++++++++++++
> tcg/s390/tcg-target.h | 48 +++++++++++++++++++++++
> 4 files changed, 157 insertions(+), 1 deletions(-)
> create mode 100644 tcg/s390/tcg-target.c
> create mode 100644 tcg/s390/tcg-target.h
>
> diff --git a/dyngen-exec.h b/dyngen-exec.h
> index 86e61c3..0353f36 100644
> --- a/dyngen-exec.h
> +++ b/dyngen-exec.h
> @@ -117,7 +117,7 @@ extern int printf(const char *, ...);
>
> /* The return address may point to the start of the next instruction.
> Subtracting one gets us the call instruction itself. */
> -#if defined(__s390__)
> +#if defined(__s390__) && !defined(__s390x__)
> # define GETPC() ((void*)(((unsigned long)__builtin_return_address(0) & 0x7fffffffUL) - 1))
> #elif defined(__arm__)
> /* Thumb return addresses have the low bit set, so we need to subtract two.
> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
> index 4e23b4a..0e222e3 100644
> --- a/target-s390x/helper.c
> +++ b/target-s390x/helper.c
> @@ -44,6 +44,11 @@ CPUS390XState *cpu_s390x_init(const char *cpu_model)
> return env;
> }
>
> +target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
> +{
> + return addr;
> +}
> +
Why does it appear in this patch? It has nothing to do with TCG support.
> void cpu_reset(CPUS390XState *env)
> {
> if (qemu_loglevel_mask(CPU_LOG_RESET)) {
> diff --git a/tcg/s390/tcg-target.c b/tcg/s390/tcg-target.c
> new file mode 100644
> index 0000000..53bbf69
> --- /dev/null
> +++ b/tcg/s390/tcg-target.c
> @@ -0,0 +1,103 @@
> +/*
> + * Tiny Code Generator for QEMU
> + *
> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> + * of this software and associated documentation files (the "Software"), to deal
> + * in the Software without restriction, including without limitation the rights
> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> + * copies of the Software, and to permit persons to whom the Software is
> + * furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> + * THE SOFTWARE.
> + */
> +
> +static const int tcg_target_reg_alloc_order[] = {
> +};
> +
> +static const int tcg_target_call_iarg_regs[] = {
> +};
> +
> +static const int tcg_target_call_oarg_regs[] = {
> +};
> +
> +static void patch_reloc(uint8_t *code_ptr, int type,
> + tcg_target_long value, tcg_target_long addend)
> +{
> + tcg_abort();
> +}
> +
> +static inline int tcg_target_get_call_iarg_regs_count(int flags)
> +{
> + tcg_abort();
> + return 0;
> +}
> +
> +/* parse target specific constraints */
> +static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
> +{
> + tcg_abort();
> + return 0;
> +}
> +
> +/* Test if a constant matches the constraint. */
> +static inline int tcg_target_const_match(tcg_target_long val,
> + const TCGArgConstraint *arg_ct)
> +{
> + tcg_abort();
> + return 0;
> +}
> +
> +/* load a register with an immediate value */
> +static inline void tcg_out_movi(TCGContext *s, TCGType type,
> + int ret, tcg_target_long arg)
> +{
> + tcg_abort();
> +}
> +
> +/* load data without address translation or endianness conversion */
> +static inline void tcg_out_ld(TCGContext *s, TCGType type, int arg,
> + int arg1, tcg_target_long arg2)
> +{
> + tcg_abort();
> +}
> +
> +static inline void tcg_out_st(TCGContext *s, TCGType type, int arg,
> + int arg1, tcg_target_long arg2)
> +{
> + tcg_abort();
> +}
> +
> +static inline void tcg_out_op(TCGContext *s, int opc,
> + const TCGArg *args, const int *const_args)
> +{
> + tcg_abort();
> +}
> +
> +void tcg_target_init(TCGContext *s)
> +{
> +}
This is probably here that tcg_abort() is actually the most important.
> +
> +void tcg_target_qemu_prologue(TCGContext *s)
> +{
> +}
> +
Also adding it here doesn't cost a lot.
> +static inline void tcg_out_mov(TCGContext *s, int ret, int arg)
> +{
> + tcg_abort();
> +}
> +
> +static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val)
> +{
> + tcg_abort();
> +}
> diff --git a/tcg/s390/tcg-target.h b/tcg/s390/tcg-target.h
> new file mode 100644
> index 0000000..2e66553
> --- /dev/null
> +++ b/tcg/s390/tcg-target.h
> @@ -0,0 +1,48 @@
> +/*
> + * Tiny Code Generator for QEMU
> + *
> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> + * of this software and associated documentation files (the "Software"), to deal
> + * in the Software without restriction, including without limitation the rights
> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> + * copies of the Software, and to permit persons to whom the Software is
> + * furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> + * THE SOFTWARE.
> + */
> +#define TCG_TARGET_S390 1
> +
> +#define TCG_TARGET_REG_BITS 64
> +#define TCG_TARGET_WORDS_BIGENDIAN
> +
> +#define TCG_TARGET_NB_REGS 0
> +
> +enum {
> + TCG_AREG0 = 0,
> +};
As this is defined in Uli's patch, I think it might be a good idea to
define that correctly. Same for the list of registers.
> +/* used for function call generation */
> +#define TCG_REG_CALL_STACK 0
> +#define TCG_TARGET_STACK_ALIGN 8
> +#define TCG_TARGET_CALL_STACK_OFFSET 0
> +
> +static inline void flush_icache_range(unsigned long start, unsigned long stop)
> +{
> +#if QEMU_GNUC_PREREQ(4, 1)
> + void __clear_cache(char *beg, char *end);
> + __clear_cache((char *) start, (char *) stop);
You should use __builtin___clear_cache instead to avoid having to
declare the function.
> +#else
> +#error not implemented
> +#endif
> +}
> --
> 1.6.0.2
>
>
>
>
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation
2009-11-30 18:18 ` Aurelien Jarno
@ 2009-11-30 22:27 ` Alexander Graf
2009-12-02 8:16 ` Aurelien Jarno
0 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-30 22:27 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> On Thu, Nov 26, 2009 at 02:23:12PM +0100, Alexander Graf wrote:
>> Qemu won't let us run a KVM target without having host TCG support. Well, for
>> now we don't have any so let's implement a fake target that only stubs out
>> everything.
>>
>> I tried to keep the patch as close to Uli's source as possible, so whenever
>> he feels like it he can easily diff his version against this one.
>
> Please find the comments below.
>
>> Signed-off-by: Alexander Graf <agraf@suse.de>
>> ---
>> dyngen-exec.h | 2 +-
>> target-s390x/helper.c | 5 ++
>> tcg/s390/tcg-target.c | 103 +++++++++++++++++++++++++++++++++++++++++++++++++
>> tcg/s390/tcg-target.h | 48 +++++++++++++++++++++++
>> 4 files changed, 157 insertions(+), 1 deletions(-)
>> create mode 100644 tcg/s390/tcg-target.c
>> create mode 100644 tcg/s390/tcg-target.h
>>
>> diff --git a/dyngen-exec.h b/dyngen-exec.h
>> index 86e61c3..0353f36 100644
>> --- a/dyngen-exec.h
>> +++ b/dyngen-exec.h
>> @@ -117,7 +117,7 @@ extern int printf(const char *, ...);
>>
>> /* The return address may point to the start of the next instruction.
>> Subtracting one gets us the call instruction itself. */
>> -#if defined(__s390__)
>> +#if defined(__s390__) && !defined(__s390x__)
>> # define GETPC() ((void*)(((unsigned long)__builtin_return_address(0) & 0x7fffffffUL) - 1))
>> #elif defined(__arm__)
>> /* Thumb return addresses have the low bit set, so we need to subtract two.
>> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
>> index 4e23b4a..0e222e3 100644
>> --- a/target-s390x/helper.c
>> +++ b/target-s390x/helper.c
>> @@ -44,6 +44,11 @@ CPUS390XState *cpu_s390x_init(const char *cpu_model)
>> return env;
>> }
>>
>> +target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
>> +{
>> + return addr;
>> +}
>> +
>
> Why does it appear in this patch? It has nothing to do with TCG support.
I don't remember. What is this used for anyways?
>> void cpu_reset(CPUS390XState *env)
>> {
>> if (qemu_loglevel_mask(CPU_LOG_RESET)) {
>> diff --git a/tcg/s390/tcg-target.c b/tcg/s390/tcg-target.c
>> new file mode 100644
>> index 0000000..53bbf69
>> --- /dev/null
>> +++ b/tcg/s390/tcg-target.c
>> @@ -0,0 +1,103 @@
>> +/*
>> + * Tiny Code Generator for QEMU
>> + *
>> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
>> + *
>> + * Permission is hereby granted, free of charge, to any person obtaining a copy
>> + * of this software and associated documentation files (the "Software"), to deal
>> + * in the Software without restriction, including without limitation the rights
>> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
>> + * copies of the Software, and to permit persons to whom the Software is
>> + * furnished to do so, subject to the following conditions:
>> + *
>> + * The above copyright notice and this permission notice shall be included in
>> + * all copies or substantial portions of the Software.
>> + *
>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
>> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
>> + * THE SOFTWARE.
>> + */
>> +
>> +static const int tcg_target_reg_alloc_order[] = {
>> +};
>> +
>> +static const int tcg_target_call_iarg_regs[] = {
>> +};
>> +
>> +static const int tcg_target_call_oarg_regs[] = {
>> +};
>> +
>> +static void patch_reloc(uint8_t *code_ptr, int type,
>> + tcg_target_long value, tcg_target_long addend)
>> +{
>> + tcg_abort();
>> +}
>> +
>> +static inline int tcg_target_get_call_iarg_regs_count(int flags)
>> +{
>> + tcg_abort();
>> + return 0;
>> +}
>> +
>> +/* parse target specific constraints */
>> +static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
>> +{
>> + tcg_abort();
>> + return 0;
>> +}
>> +
>> +/* Test if a constant matches the constraint. */
>> +static inline int tcg_target_const_match(tcg_target_long val,
>> + const TCGArgConstraint *arg_ct)
>> +{
>> + tcg_abort();
>> + return 0;
>> +}
>> +
>> +/* load a register with an immediate value */
>> +static inline void tcg_out_movi(TCGContext *s, TCGType type,
>> + int ret, tcg_target_long arg)
>> +{
>> + tcg_abort();
>> +}
>> +
>> +/* load data without address translation or endianness conversion */
>> +static inline void tcg_out_ld(TCGContext *s, TCGType type, int arg,
>> + int arg1, tcg_target_long arg2)
>> +{
>> + tcg_abort();
>> +}
>> +
>> +static inline void tcg_out_st(TCGContext *s, TCGType type, int arg,
>> + int arg1, tcg_target_long arg2)
>> +{
>> + tcg_abort();
>> +}
>> +
>> +static inline void tcg_out_op(TCGContext *s, int opc,
>> + const TCGArg *args, const int *const_args)
>> +{
>> + tcg_abort();
>> +}
>> +
>> +void tcg_target_init(TCGContext *s)
>> +{
>> +}
>
> This is probably here that tcg_abort() is actually the most important.
>
>> +
>> +void tcg_target_qemu_prologue(TCGContext *s)
>> +{
>> +}
>> +
>
> Also adding it here doesn't cost a lot.
Both places get called with KVM as well. If I call tcg_abort() here KVM doesn't start.
>
>> +static inline void tcg_out_mov(TCGContext *s, int ret, int arg)
>> +{
>> + tcg_abort();
>> +}
>> +
>> +static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val)
>> +{
>> + tcg_abort();
>> +}
>> diff --git a/tcg/s390/tcg-target.h b/tcg/s390/tcg-target.h
>> new file mode 100644
>> index 0000000..2e66553
>> --- /dev/null
>> +++ b/tcg/s390/tcg-target.h
>> @@ -0,0 +1,48 @@
>> +/*
>> + * Tiny Code Generator for QEMU
>> + *
>> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
>> + *
>> + * Permission is hereby granted, free of charge, to any person obtaining a copy
>> + * of this software and associated documentation files (the "Software"), to deal
>> + * in the Software without restriction, including without limitation the rights
>> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
>> + * copies of the Software, and to permit persons to whom the Software is
>> + * furnished to do so, subject to the following conditions:
>> + *
>> + * The above copyright notice and this permission notice shall be included in
>> + * all copies or substantial portions of the Software.
>> + *
>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
>> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
>> + * THE SOFTWARE.
>> + */
>> +#define TCG_TARGET_S390 1
>> +
>> +#define TCG_TARGET_REG_BITS 64
>> +#define TCG_TARGET_WORDS_BIGENDIAN
>> +
>> +#define TCG_TARGET_NB_REGS 0
>> +
>> +enum {
>> + TCG_AREG0 = 0,
>> +};
>
> As this is defined in Uli's patch, I think it might be a good idea to
> define that correctly. Same for the list of registers.
I wasn't sure how much of the real architecture I should actually keep. So you think registers belong in?
>> +/* used for function call generation */
>> +#define TCG_REG_CALL_STACK 0
>> +#define TCG_TARGET_STACK_ALIGN 8
>> +#define TCG_TARGET_CALL_STACK_OFFSET 0
>> +
>> +static inline void flush_icache_range(unsigned long start, unsigned long stop)
>> +{
>> +#if QEMU_GNUC_PREREQ(4, 1)
>> + void __clear_cache(char *beg, char *end);
>> + __clear_cache((char *) start, (char *) stop);
>
> You should use __builtin___clear_cache instead to avoid having to
> declare the function.
Oh, ok :)
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation
2009-11-30 22:27 ` Alexander Graf
@ 2009-12-02 8:16 ` Aurelien Jarno
2009-12-02 8:29 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-12-02 8:16 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Mon, Nov 30, 2009 at 11:27:53PM +0100, Alexander Graf wrote:
>
> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>
> > On Thu, Nov 26, 2009 at 02:23:12PM +0100, Alexander Graf wrote:
> >> Qemu won't let us run a KVM target without having host TCG support. Well, for
> >> now we don't have any so let's implement a fake target that only stubs out
> >> everything.
> >>
> >> I tried to keep the patch as close to Uli's source as possible, so whenever
> >> he feels like it he can easily diff his version against this one.
> >
> > Please find the comments below.
> >
> >> Signed-off-by: Alexander Graf <agraf@suse.de>
> >> ---
> >> dyngen-exec.h | 2 +-
> >> target-s390x/helper.c | 5 ++
> >> tcg/s390/tcg-target.c | 103 +++++++++++++++++++++++++++++++++++++++++++++++++
> >> tcg/s390/tcg-target.h | 48 +++++++++++++++++++++++
> >> 4 files changed, 157 insertions(+), 1 deletions(-)
> >> create mode 100644 tcg/s390/tcg-target.c
> >> create mode 100644 tcg/s390/tcg-target.h
> >>
> >> diff --git a/dyngen-exec.h b/dyngen-exec.h
> >> index 86e61c3..0353f36 100644
> >> --- a/dyngen-exec.h
> >> +++ b/dyngen-exec.h
> >> @@ -117,7 +117,7 @@ extern int printf(const char *, ...);
> >>
> >> /* The return address may point to the start of the next instruction.
> >> Subtracting one gets us the call instruction itself. */
> >> -#if defined(__s390__)
> >> +#if defined(__s390__) && !defined(__s390x__)
> >> # define GETPC() ((void*)(((unsigned long)__builtin_return_address(0) & 0x7fffffffUL) - 1))
> >> #elif defined(__arm__)
> >> /* Thumb return addresses have the low bit set, so we need to subtract two.
> >> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
> >> index 4e23b4a..0e222e3 100644
> >> --- a/target-s390x/helper.c
> >> +++ b/target-s390x/helper.c
> >> @@ -44,6 +44,11 @@ CPUS390XState *cpu_s390x_init(const char *cpu_model)
> >> return env;
> >> }
> >>
> >> +target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
> >> +{
> >> + return addr;
> >> +}
> >> +
> >
> > Why does it appear in this patch? It has nothing to do with TCG support.
>
> I don't remember. What is this used for anyways?
If it is not used, maybe it's better to remove it.
> >> void cpu_reset(CPUS390XState *env)
> >> {
> >> if (qemu_loglevel_mask(CPU_LOG_RESET)) {
> >> diff --git a/tcg/s390/tcg-target.c b/tcg/s390/tcg-target.c
> >> new file mode 100644
> >> index 0000000..53bbf69
> >> --- /dev/null
> >> +++ b/tcg/s390/tcg-target.c
> >> @@ -0,0 +1,103 @@
> >> +/*
> >> + * Tiny Code Generator for QEMU
> >> + *
> >> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
> >> + *
> >> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> >> + * of this software and associated documentation files (the "Software"), to deal
> >> + * in the Software without restriction, including without limitation the rights
> >> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> >> + * copies of the Software, and to permit persons to whom the Software is
> >> + * furnished to do so, subject to the following conditions:
> >> + *
> >> + * The above copyright notice and this permission notice shall be included in
> >> + * all copies or substantial portions of the Software.
> >> + *
> >> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> >> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> >> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> >> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> >> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> >> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> >> + * THE SOFTWARE.
> >> + */
> >> +
> >> +static const int tcg_target_reg_alloc_order[] = {
> >> +};
> >> +
> >> +static const int tcg_target_call_iarg_regs[] = {
> >> +};
> >> +
> >> +static const int tcg_target_call_oarg_regs[] = {
> >> +};
> >> +
> >> +static void patch_reloc(uint8_t *code_ptr, int type,
> >> + tcg_target_long value, tcg_target_long addend)
> >> +{
> >> + tcg_abort();
> >> +}
> >> +
> >> +static inline int tcg_target_get_call_iarg_regs_count(int flags)
> >> +{
> >> + tcg_abort();
> >> + return 0;
> >> +}
> >> +
> >> +/* parse target specific constraints */
> >> +static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
> >> +{
> >> + tcg_abort();
> >> + return 0;
> >> +}
> >> +
> >> +/* Test if a constant matches the constraint. */
> >> +static inline int tcg_target_const_match(tcg_target_long val,
> >> + const TCGArgConstraint *arg_ct)
> >> +{
> >> + tcg_abort();
> >> + return 0;
> >> +}
> >> +
> >> +/* load a register with an immediate value */
> >> +static inline void tcg_out_movi(TCGContext *s, TCGType type,
> >> + int ret, tcg_target_long arg)
> >> +{
> >> + tcg_abort();
> >> +}
> >> +
> >> +/* load data without address translation or endianness conversion */
> >> +static inline void tcg_out_ld(TCGContext *s, TCGType type, int arg,
> >> + int arg1, tcg_target_long arg2)
> >> +{
> >> + tcg_abort();
> >> +}
> >> +
> >> +static inline void tcg_out_st(TCGContext *s, TCGType type, int arg,
> >> + int arg1, tcg_target_long arg2)
> >> +{
> >> + tcg_abort();
> >> +}
> >> +
> >> +static inline void tcg_out_op(TCGContext *s, int opc,
> >> + const TCGArg *args, const int *const_args)
> >> +{
> >> + tcg_abort();
> >> +}
> >> +
> >> +void tcg_target_init(TCGContext *s)
> >> +{
> >> +}
> >
> > This is probably here that tcg_abort() is actually the most important.
> >
> >> +
> >> +void tcg_target_qemu_prologue(TCGContext *s)
> >> +{
> >> +}
> >> +
> >
> > Also adding it here doesn't cost a lot.
>
> Both places get called with KVM as well. If I call tcg_abort() here KVM doesn't start.
Then can you add a comment explaining that code is missing.
I am very reluctant in introducing missing/wrong code in qemu.
Experience has shown that it stays for a very long time, until someone
spend hours or days trying to debug an issue.
> >
> >> +static inline void tcg_out_mov(TCGContext *s, int ret, int arg)
> >> +{
> >> + tcg_abort();
> >> +}
> >> +
> >> +static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val)
> >> +{
> >> + tcg_abort();
> >> +}
> >> diff --git a/tcg/s390/tcg-target.h b/tcg/s390/tcg-target.h
> >> new file mode 100644
> >> index 0000000..2e66553
> >> --- /dev/null
> >> +++ b/tcg/s390/tcg-target.h
> >> @@ -0,0 +1,48 @@
> >> +/*
> >> + * Tiny Code Generator for QEMU
> >> + *
> >> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
> >> + *
> >> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> >> + * of this software and associated documentation files (the "Software"), to deal
> >> + * in the Software without restriction, including without limitation the rights
> >> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> >> + * copies of the Software, and to permit persons to whom the Software is
> >> + * furnished to do so, subject to the following conditions:
> >> + *
> >> + * The above copyright notice and this permission notice shall be included in
> >> + * all copies or substantial portions of the Software.
> >> + *
> >> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> >> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> >> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> >> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> >> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> >> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> >> + * THE SOFTWARE.
> >> + */
> >> +#define TCG_TARGET_S390 1
> >> +
> >> +#define TCG_TARGET_REG_BITS 64
> >> +#define TCG_TARGET_WORDS_BIGENDIAN
> >> +
> >> +#define TCG_TARGET_NB_REGS 0
> >> +
> >> +enum {
> >> + TCG_AREG0 = 0,
> >> +};
> >
> > As this is defined in Uli's patch, I think it might be a good idea to
> > define that correctly. Same for the list of registers.
>
> I wasn't sure how much of the real architecture I should actually keep. So you think registers belong in?
Again what chokes me is that it introduces wrong code. Especially
TCG_AREG0 = 0. Introducing the registers will fix that.
> >> +/* used for function call generation */
> >> +#define TCG_REG_CALL_STACK 0
> >> +#define TCG_TARGET_STACK_ALIGN 8
> >> +#define TCG_TARGET_CALL_STACK_OFFSET 0
> >> +
> >> +static inline void flush_icache_range(unsigned long start, unsigned long stop)
> >> +{
> >> +#if QEMU_GNUC_PREREQ(4, 1)
> >> + void __clear_cache(char *beg, char *end);
> >> + __clear_cache((char *) start, (char *) stop);
> >
> > You should use __builtin___clear_cache instead to avoid having to
> > declare the function.
>
> Oh, ok :)
>
> Alex
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation
2009-12-02 8:16 ` Aurelien Jarno
@ 2009-12-02 8:29 ` Alexander Graf
2009-12-02 8:41 ` Aurelien Jarno
0 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-12-02 8:29 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 02.12.2009, at 09:16, Aurelien Jarno wrote:
> On Mon, Nov 30, 2009 at 11:27:53PM +0100, Alexander Graf wrote:
>>
>> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>>
>>> On Thu, Nov 26, 2009 at 02:23:12PM +0100, Alexander Graf wrote:
>>>> Qemu won't let us run a KVM target without having host TCG support. Well, for
>>>> now we don't have any so let's implement a fake target that only stubs out
>>>> everything.
>>>>
>>>> I tried to keep the patch as close to Uli's source as possible, so whenever
>>>> he feels like it he can easily diff his version against this one.
>>>
>>> Please find the comments below.
>>>
>>>> Signed-off-by: Alexander Graf <agraf@suse.de>
>>>> ---
>>>> dyngen-exec.h | 2 +-
>>>> target-s390x/helper.c | 5 ++
>>>> tcg/s390/tcg-target.c | 103 +++++++++++++++++++++++++++++++++++++++++++++++++
>>>> tcg/s390/tcg-target.h | 48 +++++++++++++++++++++++
>>>> 4 files changed, 157 insertions(+), 1 deletions(-)
>>>> create mode 100644 tcg/s390/tcg-target.c
>>>> create mode 100644 tcg/s390/tcg-target.h
>>>>
>>>> diff --git a/dyngen-exec.h b/dyngen-exec.h
>>>> index 86e61c3..0353f36 100644
>>>> --- a/dyngen-exec.h
>>>> +++ b/dyngen-exec.h
>>>> @@ -117,7 +117,7 @@ extern int printf(const char *, ...);
>>>>
>>>> /* The return address may point to the start of the next instruction.
>>>> Subtracting one gets us the call instruction itself. */
>>>> -#if defined(__s390__)
>>>> +#if defined(__s390__) && !defined(__s390x__)
>>>> # define GETPC() ((void*)(((unsigned long)__builtin_return_address(0) & 0x7fffffffUL) - 1))
>>>> #elif defined(__arm__)
>>>> /* Thumb return addresses have the low bit set, so we need to subtract two.
>>>> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
>>>> index 4e23b4a..0e222e3 100644
>>>> --- a/target-s390x/helper.c
>>>> +++ b/target-s390x/helper.c
>>>> @@ -44,6 +44,11 @@ CPUS390XState *cpu_s390x_init(const char *cpu_model)
>>>> return env;
>>>> }
>>>>
>>>> +target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
>>>> +{
>>>> + return addr;
>>>> +}
>>>> +
>>>
>>> Why does it appear in this patch? It has nothing to do with TCG support.
>>
>> I don't remember. What is this used for anyways?
>
> If it is not used, maybe it's better to remove it.
It's called from exec.c.
>
>>>> void cpu_reset(CPUS390XState *env)
>>>> {
>>>> if (qemu_loglevel_mask(CPU_LOG_RESET)) {
>>>> diff --git a/tcg/s390/tcg-target.c b/tcg/s390/tcg-target.c
>>>> new file mode 100644
>>>> index 0000000..53bbf69
>>>> --- /dev/null
>>>> +++ b/tcg/s390/tcg-target.c
>>>> @@ -0,0 +1,103 @@
>>>> +/*
>>>> + * Tiny Code Generator for QEMU
>>>> + *
>>>> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
>>>> + *
>>>> + * Permission is hereby granted, free of charge, to any person obtaining a copy
>>>> + * of this software and associated documentation files (the "Software"), to deal
>>>> + * in the Software without restriction, including without limitation the rights
>>>> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
>>>> + * copies of the Software, and to permit persons to whom the Software is
>>>> + * furnished to do so, subject to the following conditions:
>>>> + *
>>>> + * The above copyright notice and this permission notice shall be included in
>>>> + * all copies or substantial portions of the Software.
>>>> + *
>>>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
>>>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
>>>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
>>>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
>>>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
>>>> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
>>>> + * THE SOFTWARE.
>>>> + */
>>>> +
>>>> +static const int tcg_target_reg_alloc_order[] = {
>>>> +};
>>>> +
>>>> +static const int tcg_target_call_iarg_regs[] = {
>>>> +};
>>>> +
>>>> +static const int tcg_target_call_oarg_regs[] = {
>>>> +};
>>>> +
>>>> +static void patch_reloc(uint8_t *code_ptr, int type,
>>>> + tcg_target_long value, tcg_target_long addend)
>>>> +{
>>>> + tcg_abort();
>>>> +}
>>>> +
>>>> +static inline int tcg_target_get_call_iarg_regs_count(int flags)
>>>> +{
>>>> + tcg_abort();
>>>> + return 0;
>>>> +}
>>>> +
>>>> +/* parse target specific constraints */
>>>> +static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
>>>> +{
>>>> + tcg_abort();
>>>> + return 0;
>>>> +}
>>>> +
>>>> +/* Test if a constant matches the constraint. */
>>>> +static inline int tcg_target_const_match(tcg_target_long val,
>>>> + const TCGArgConstraint *arg_ct)
>>>> +{
>>>> + tcg_abort();
>>>> + return 0;
>>>> +}
>>>> +
>>>> +/* load a register with an immediate value */
>>>> +static inline void tcg_out_movi(TCGContext *s, TCGType type,
>>>> + int ret, tcg_target_long arg)
>>>> +{
>>>> + tcg_abort();
>>>> +}
>>>> +
>>>> +/* load data without address translation or endianness conversion */
>>>> +static inline void tcg_out_ld(TCGContext *s, TCGType type, int arg,
>>>> + int arg1, tcg_target_long arg2)
>>>> +{
>>>> + tcg_abort();
>>>> +}
>>>> +
>>>> +static inline void tcg_out_st(TCGContext *s, TCGType type, int arg,
>>>> + int arg1, tcg_target_long arg2)
>>>> +{
>>>> + tcg_abort();
>>>> +}
>>>> +
>>>> +static inline void tcg_out_op(TCGContext *s, int opc,
>>>> + const TCGArg *args, const int *const_args)
>>>> +{
>>>> + tcg_abort();
>>>> +}
>>>> +
>>>> +void tcg_target_init(TCGContext *s)
>>>> +{
>>>> +}
>>>
>>> This is probably here that tcg_abort() is actually the most important.
>>>
>>>> +
>>>> +void tcg_target_qemu_prologue(TCGContext *s)
>>>> +{
>>>> +}
>>>> +
>>>
>>> Also adding it here doesn't cost a lot.
>>
>> Both places get called with KVM as well. If I call tcg_abort() here KVM doesn't start.
>
> Then can you add a comment explaining that code is missing.
>
> I am very reluctant in introducing missing/wrong code in qemu.
> Experience has shown that it stays for a very long time, until someone
> spend hours or days trying to debug an issue.
Alright.
>
>>>
>>>> +static inline void tcg_out_mov(TCGContext *s, int ret, int arg)
>>>> +{
>>>> + tcg_abort();
>>>> +}
>>>> +
>>>> +static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val)
>>>> +{
>>>> + tcg_abort();
>>>> +}
>>>> diff --git a/tcg/s390/tcg-target.h b/tcg/s390/tcg-target.h
>>>> new file mode 100644
>>>> index 0000000..2e66553
>>>> --- /dev/null
>>>> +++ b/tcg/s390/tcg-target.h
>>>> @@ -0,0 +1,48 @@
>>>> +/*
>>>> + * Tiny Code Generator for QEMU
>>>> + *
>>>> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
>>>> + *
>>>> + * Permission is hereby granted, free of charge, to any person obtaining a copy
>>>> + * of this software and associated documentation files (the "Software"), to deal
>>>> + * in the Software without restriction, including without limitation the rights
>>>> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
>>>> + * copies of the Software, and to permit persons to whom the Software is
>>>> + * furnished to do so, subject to the following conditions:
>>>> + *
>>>> + * The above copyright notice and this permission notice shall be included in
>>>> + * all copies or substantial portions of the Software.
>>>> + *
>>>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
>>>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
>>>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
>>>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
>>>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
>>>> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
>>>> + * THE SOFTWARE.
>>>> + */
>>>> +#define TCG_TARGET_S390 1
>>>> +
>>>> +#define TCG_TARGET_REG_BITS 64
>>>> +#define TCG_TARGET_WORDS_BIGENDIAN
>>>> +
>>>> +#define TCG_TARGET_NB_REGS 0
>>>> +
>>>> +enum {
>>>> + TCG_AREG0 = 0,
>>>> +};
>>>
>>> As this is defined in Uli's patch, I think it might be a good idea to
>>> define that correctly. Same for the list of registers.
>>
>> I wasn't sure how much of the real architecture I should actually keep. So you think registers belong in?
>
> Again what chokes me is that it introduces wrong code. Especially
> TCG_AREG0 = 0. Introducing the registers will fix that.
Ok.
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation
2009-12-02 8:29 ` Alexander Graf
@ 2009-12-02 8:41 ` Aurelien Jarno
2009-12-02 8:44 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-12-02 8:41 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Wed, Dec 02, 2009 at 09:29:59AM +0100, Alexander Graf wrote:
>
> On 02.12.2009, at 09:16, Aurelien Jarno wrote:
>
> > On Mon, Nov 30, 2009 at 11:27:53PM +0100, Alexander Graf wrote:
> >>
> >> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> >>
> >>> On Thu, Nov 26, 2009 at 02:23:12PM +0100, Alexander Graf wrote:
> >>>> Qemu won't let us run a KVM target without having host TCG support. Well, for
> >>>> now we don't have any so let's implement a fake target that only stubs out
> >>>> everything.
> >>>>
> >>>> I tried to keep the patch as close to Uli's source as possible, so whenever
> >>>> he feels like it he can easily diff his version against this one.
> >>>
> >>> Please find the comments below.
> >>>
> >>>> Signed-off-by: Alexander Graf <agraf@suse.de>
> >>>> ---
> >>>> dyngen-exec.h | 2 +-
> >>>> target-s390x/helper.c | 5 ++
> >>>> tcg/s390/tcg-target.c | 103 +++++++++++++++++++++++++++++++++++++++++++++++++
> >>>> tcg/s390/tcg-target.h | 48 +++++++++++++++++++++++
> >>>> 4 files changed, 157 insertions(+), 1 deletions(-)
> >>>> create mode 100644 tcg/s390/tcg-target.c
> >>>> create mode 100644 tcg/s390/tcg-target.h
> >>>>
> >>>> diff --git a/dyngen-exec.h b/dyngen-exec.h
> >>>> index 86e61c3..0353f36 100644
> >>>> --- a/dyngen-exec.h
> >>>> +++ b/dyngen-exec.h
> >>>> @@ -117,7 +117,7 @@ extern int printf(const char *, ...);
> >>>>
> >>>> /* The return address may point to the start of the next instruction.
> >>>> Subtracting one gets us the call instruction itself. */
> >>>> -#if defined(__s390__)
> >>>> +#if defined(__s390__) && !defined(__s390x__)
> >>>> # define GETPC() ((void*)(((unsigned long)__builtin_return_address(0) & 0x7fffffffUL) - 1))
> >>>> #elif defined(__arm__)
> >>>> /* Thumb return addresses have the low bit set, so we need to subtract two.
> >>>> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
> >>>> index 4e23b4a..0e222e3 100644
> >>>> --- a/target-s390x/helper.c
> >>>> +++ b/target-s390x/helper.c
> >>>> @@ -44,6 +44,11 @@ CPUS390XState *cpu_s390x_init(const char *cpu_model)
> >>>> return env;
> >>>> }
> >>>>
> >>>> +target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
> >>>> +{
> >>>> + return addr;
> >>>> +}
> >>>> +
> >>>
> >>> Why does it appear in this patch? It has nothing to do with TCG support.
> >>
> >> I don't remember. What is this used for anyways?
> >
> > If it is not used, maybe it's better to remove it.
>
> It's called from exec.c.
Then it's clearly not part of this patch, it should probably be part of
patch 1 or 6 instead.
> >
> >>>> void cpu_reset(CPUS390XState *env)
> >>>> {
> >>>> if (qemu_loglevel_mask(CPU_LOG_RESET)) {
> >>>> diff --git a/tcg/s390/tcg-target.c b/tcg/s390/tcg-target.c
> >>>> new file mode 100644
> >>>> index 0000000..53bbf69
> >>>> --- /dev/null
> >>>> +++ b/tcg/s390/tcg-target.c
> >>>> @@ -0,0 +1,103 @@
> >>>> +/*
> >>>> + * Tiny Code Generator for QEMU
> >>>> + *
> >>>> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
> >>>> + *
> >>>> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> >>>> + * of this software and associated documentation files (the "Software"), to deal
> >>>> + * in the Software without restriction, including without limitation the rights
> >>>> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> >>>> + * copies of the Software, and to permit persons to whom the Software is
> >>>> + * furnished to do so, subject to the following conditions:
> >>>> + *
> >>>> + * The above copyright notice and this permission notice shall be included in
> >>>> + * all copies or substantial portions of the Software.
> >>>> + *
> >>>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> >>>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> >>>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> >>>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> >>>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> >>>> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> >>>> + * THE SOFTWARE.
> >>>> + */
> >>>> +
> >>>> +static const int tcg_target_reg_alloc_order[] = {
> >>>> +};
> >>>> +
> >>>> +static const int tcg_target_call_iarg_regs[] = {
> >>>> +};
> >>>> +
> >>>> +static const int tcg_target_call_oarg_regs[] = {
> >>>> +};
> >>>> +
> >>>> +static void patch_reloc(uint8_t *code_ptr, int type,
> >>>> + tcg_target_long value, tcg_target_long addend)
> >>>> +{
> >>>> + tcg_abort();
> >>>> +}
> >>>> +
> >>>> +static inline int tcg_target_get_call_iarg_regs_count(int flags)
> >>>> +{
> >>>> + tcg_abort();
> >>>> + return 0;
> >>>> +}
> >>>> +
> >>>> +/* parse target specific constraints */
> >>>> +static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
> >>>> +{
> >>>> + tcg_abort();
> >>>> + return 0;
> >>>> +}
> >>>> +
> >>>> +/* Test if a constant matches the constraint. */
> >>>> +static inline int tcg_target_const_match(tcg_target_long val,
> >>>> + const TCGArgConstraint *arg_ct)
> >>>> +{
> >>>> + tcg_abort();
> >>>> + return 0;
> >>>> +}
> >>>> +
> >>>> +/* load a register with an immediate value */
> >>>> +static inline void tcg_out_movi(TCGContext *s, TCGType type,
> >>>> + int ret, tcg_target_long arg)
> >>>> +{
> >>>> + tcg_abort();
> >>>> +}
> >>>> +
> >>>> +/* load data without address translation or endianness conversion */
> >>>> +static inline void tcg_out_ld(TCGContext *s, TCGType type, int arg,
> >>>> + int arg1, tcg_target_long arg2)
> >>>> +{
> >>>> + tcg_abort();
> >>>> +}
> >>>> +
> >>>> +static inline void tcg_out_st(TCGContext *s, TCGType type, int arg,
> >>>> + int arg1, tcg_target_long arg2)
> >>>> +{
> >>>> + tcg_abort();
> >>>> +}
> >>>> +
> >>>> +static inline void tcg_out_op(TCGContext *s, int opc,
> >>>> + const TCGArg *args, const int *const_args)
> >>>> +{
> >>>> + tcg_abort();
> >>>> +}
> >>>> +
> >>>> +void tcg_target_init(TCGContext *s)
> >>>> +{
> >>>> +}
> >>>
> >>> This is probably here that tcg_abort() is actually the most important.
> >>>
> >>>> +
> >>>> +void tcg_target_qemu_prologue(TCGContext *s)
> >>>> +{
> >>>> +}
> >>>> +
> >>>
> >>> Also adding it here doesn't cost a lot.
> >>
> >> Both places get called with KVM as well. If I call tcg_abort() here KVM doesn't start.
> >
> > Then can you add a comment explaining that code is missing.
> >
> > I am very reluctant in introducing missing/wrong code in qemu.
> > Experience has shown that it stays for a very long time, until someone
> > spend hours or days trying to debug an issue.
>
> Alright.
>
> >
> >>>
> >>>> +static inline void tcg_out_mov(TCGContext *s, int ret, int arg)
> >>>> +{
> >>>> + tcg_abort();
> >>>> +}
> >>>> +
> >>>> +static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val)
> >>>> +{
> >>>> + tcg_abort();
> >>>> +}
> >>>> diff --git a/tcg/s390/tcg-target.h b/tcg/s390/tcg-target.h
> >>>> new file mode 100644
> >>>> index 0000000..2e66553
> >>>> --- /dev/null
> >>>> +++ b/tcg/s390/tcg-target.h
> >>>> @@ -0,0 +1,48 @@
> >>>> +/*
> >>>> + * Tiny Code Generator for QEMU
> >>>> + *
> >>>> + * Copyright (c) 2009 Ulrich Hecht <uli@suse.de>
> >>>> + *
> >>>> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> >>>> + * of this software and associated documentation files (the "Software"), to deal
> >>>> + * in the Software without restriction, including without limitation the rights
> >>>> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> >>>> + * copies of the Software, and to permit persons to whom the Software is
> >>>> + * furnished to do so, subject to the following conditions:
> >>>> + *
> >>>> + * The above copyright notice and this permission notice shall be included in
> >>>> + * all copies or substantial portions of the Software.
> >>>> + *
> >>>> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> >>>> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> >>>> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
> >>>> + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> >>>> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> >>>> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> >>>> + * THE SOFTWARE.
> >>>> + */
> >>>> +#define TCG_TARGET_S390 1
> >>>> +
> >>>> +#define TCG_TARGET_REG_BITS 64
> >>>> +#define TCG_TARGET_WORDS_BIGENDIAN
> >>>> +
> >>>> +#define TCG_TARGET_NB_REGS 0
> >>>> +
> >>>> +enum {
> >>>> + TCG_AREG0 = 0,
> >>>> +};
> >>>
> >>> As this is defined in Uli's patch, I think it might be a good idea to
> >>> define that correctly. Same for the list of registers.
> >>
> >> I wasn't sure how much of the real architecture I should actually keep. So you think registers belong in?
> >
> > Again what chokes me is that it introduces wrong code. Especially
> > TCG_AREG0 = 0. Introducing the registers will fix that.
>
> Ok.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation
2009-12-02 8:41 ` Aurelien Jarno
@ 2009-12-02 8:44 ` Alexander Graf
0 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-12-02 8:44 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 02.12.2009, at 09:41, Aurelien Jarno wrote:
> On Wed, Dec 02, 2009 at 09:29:59AM +0100, Alexander Graf wrote:
>>
>> On 02.12.2009, at 09:16, Aurelien Jarno wrote:
>>
>>> On Mon, Nov 30, 2009 at 11:27:53PM +0100, Alexander Graf wrote:
>>>>
>>>> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>>>>
>>>>> On Thu, Nov 26, 2009 at 02:23:12PM +0100, Alexander Graf wrote:
>>>>>> Qemu won't let us run a KVM target without having host TCG support. Well, for
>>>>>> now we don't have any so let's implement a fake target that only stubs out
>>>>>> everything.
>>>>>>
>>>>>> I tried to keep the patch as close to Uli's source as possible, so whenever
>>>>>> he feels like it he can easily diff his version against this one.
>>>>>
>>>>> Please find the comments below.
>>>>>
>>>>>> Signed-off-by: Alexander Graf <agraf@suse.de>
>>>>>> ---
>>>>>> dyngen-exec.h | 2 +-
>>>>>> target-s390x/helper.c | 5 ++
>>>>>> tcg/s390/tcg-target.c | 103 +++++++++++++++++++++++++++++++++++++++++++++++++
>>>>>> tcg/s390/tcg-target.h | 48 +++++++++++++++++++++++
>>>>>> 4 files changed, 157 insertions(+), 1 deletions(-)
>>>>>> create mode 100644 tcg/s390/tcg-target.c
>>>>>> create mode 100644 tcg/s390/tcg-target.h
>>>>>>
>>>>>> diff --git a/dyngen-exec.h b/dyngen-exec.h
>>>>>> index 86e61c3..0353f36 100644
>>>>>> --- a/dyngen-exec.h
>>>>>> +++ b/dyngen-exec.h
>>>>>> @@ -117,7 +117,7 @@ extern int printf(const char *, ...);
>>>>>>
>>>>>> /* The return address may point to the start of the next instruction.
>>>>>> Subtracting one gets us the call instruction itself. */
>>>>>> -#if defined(__s390__)
>>>>>> +#if defined(__s390__) && !defined(__s390x__)
>>>>>> # define GETPC() ((void*)(((unsigned long)__builtin_return_address(0) & 0x7fffffffUL) - 1))
>>>>>> #elif defined(__arm__)
>>>>>> /* Thumb return addresses have the low bit set, so we need to subtract two.
>>>>>> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
>>>>>> index 4e23b4a..0e222e3 100644
>>>>>> --- a/target-s390x/helper.c
>>>>>> +++ b/target-s390x/helper.c
>>>>>> @@ -44,6 +44,11 @@ CPUS390XState *cpu_s390x_init(const char *cpu_model)
>>>>>> return env;
>>>>>> }
>>>>>>
>>>>>> +target_phys_addr_t cpu_get_phys_page_debug(CPUState *env, target_ulong addr)
>>>>>> +{
>>>>>> + return addr;
>>>>>> +}
>>>>>> +
>>>>>
>>>>> Why does it appear in this patch? It has nothing to do with TCG support.
>>>>
>>>> I don't remember. What is this used for anyways?
>>>
>>> If it is not used, maybe it's better to remove it.
>>
>> It's called from exec.c.
>
> Then it's clearly not part of this patch, it should probably be part of
> patch 1 or 6 instead.
Yes, moved it already.
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread
* [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (2 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 03/11] S/390 fake TCG implementation Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-30 18:18 ` Aurelien Jarno
2009-11-26 13:23 ` [Qemu-devel] [PATCH 05/11] Allocate physical memory in low virtual address space Alexander Graf
` (7 subsequent siblings)
11 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
S390x was one of the first platforms that received support for KVM back in the
day. Unfortunately until now there hasn't been a qemu implementation that would
enable users to actually run guests.
So let's include support for KVM S390x in qemu!
Signed-off-by: Alexander Graf <agraf@suse.de>
---
configure | 4 +-
target-s390x/kvm.c | 463 ++++++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 466 insertions(+), 1 deletions(-)
create mode 100644 target-s390x/kvm.c
diff --git a/configure b/configure
index b616e1a..cf466ec 100755
--- a/configure
+++ b/configure
@@ -1348,6 +1348,8 @@ EOF
kvm_cflags="$kvm_cflags -I$kerneldir/arch/x86/include"
elif test "$cpu" = "ppc" -a -d "$kerneldir/arch/powerpc/include" ; then
kvm_cflags="$kvm_cflags -I$kerneldir/arch/powerpc/include"
+ elif test "$cpu" = "s390x" -a -d "$kerneldir/arch/s390/include" ; then
+ kvm_cflags="$kvm_cflags -I$kerneldir/arch/s390/include"
elif test -d "$kerneldir/arch/$cpu/include" ; then
kvm_cflags="$kvm_cflags -I$kerneldir/arch/$cpu/include"
fi
@@ -2360,7 +2362,7 @@ case "$target_arch2" in
fi
esac
case "$target_arch2" in
- i386|x86_64|ppcemb|ppc|ppc64)
+ i386|x86_64|ppcemb|ppc|ppc64|s390x)
# Make sure the target and host cpus are compatible
if test "$kvm" = "yes" -a "$target_softmmu" = "yes" -a \
\( "$target_arch2" = "$cpu" -o \
diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
new file mode 100644
index 0000000..d477664
--- /dev/null
+++ b/target-s390x/kvm.c
@@ -0,0 +1,463 @@
+/*
+ * QEMU S390x KVM implementation
+ *
+ * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <sys/types.h>
+#include <sys/ioctl.h>
+#include <sys/mman.h>
+
+#include <linux/kvm.h>
+#include <asm/ptrace.h>
+
+#include "qemu-common.h"
+#include "qemu-timer.h"
+#include "sysemu.h"
+#include "kvm.h"
+#include "cpu.h"
+#include "device_tree.h"
+
+/* #define DEBUG_KVM */
+
+#ifdef DEBUG_KVM
+#define dprintf(fmt, ...) \
+ do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
+#else
+#define dprintf(fmt, ...) \
+ do { } while (0)
+#endif
+
+#define IPA0_DIAG 0x8300
+#define IPA0_SIGP 0xae00
+#define IPA0_PRIV 0xb200
+
+#define PRIV_SCLP_CALL 0x20
+#define DIAG_KVM_HYPERCALL 0x500
+#define DIAG_KVM_BREAKPOINT 0x501
+
+#define SCP_LENGTH 0x00
+#define SCP_FUNCTION_CODE 0x02
+#define SCP_CONTROL_MASK 0x03
+#define SCP_RESPONSE_CODE 0x06
+#define SCP_MEM_CODE 0x08
+#define SCP_INCREMENT 0x0a
+
+#define ICPT_INSTRUCTION 0x04
+#define ICPT_WAITPSW 0x1c
+#define ICPT_SOFT_INTERCEPT 0x24
+#define ICPT_CPU_STOP 0x28
+#define ICPT_IO 0x40
+
+#define SIGP_RESTART 0x06
+#define SIGP_INITIAL_CPU_RESET 0x0b
+#define SIGP_STORE_STATUS_ADDR 0x0e
+#define SIGP_SET_ARCH 0x12
+
+
+int kvm_arch_init(KVMState *s, int smp_cpus)
+{
+ return 0;
+}
+
+int kvm_arch_init_vcpu(CPUState *env)
+{
+ int ret = 0;
+
+ if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL) < 0)
+ perror("cannot init reset vcpu");
+
+ return ret;
+}
+
+void kvm_arch_reset_vcpu(CPUState *env)
+{
+}
+
+int kvm_arch_put_registers(CPUState *env)
+{
+ struct kvm_regs regs;
+ int ret;
+ int i;
+
+ ret = kvm_vcpu_ioctl(env, KVM_GET_REGS, ®s);
+ if (ret < 0)
+ return ret;
+
+ for (i = 0; i < 16; i++)
+ regs.gprs[i] = env->regs[i];
+
+ ret = kvm_vcpu_ioctl(env, KVM_SET_REGS, ®s);
+ if (ret < 0)
+ return ret;
+
+ env->kvm_run->psw_addr = env->psw.addr;
+ env->kvm_run->psw_mask = env->psw.mask;
+
+ return ret;
+}
+
+int kvm_arch_get_registers(CPUState *env)
+{
+ uint32_t ret;
+ struct kvm_regs regs;
+ int i;
+
+ ret = kvm_vcpu_ioctl(env, KVM_GET_REGS, ®s);
+ if (ret < 0)
+ return ret;
+
+ for (i = 0; i < 16; i++)
+ env->regs[i] = regs.gprs[i];
+
+ env->psw.addr = env->kvm_run->psw_addr;
+ env->psw.mask = env->kvm_run->psw_mask;
+
+ return 0;
+}
+
+int kvm_arch_insert_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
+{
+ static const uint8_t diag_501[] = {0x83, 0x24, 0x05, 0x01};
+
+ if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 4, 0) ||
+ cpu_memory_rw_debug(env, bp->pc, (uint8_t *)diag_501, 4, 1))
+ return -EINVAL;
+ return 0;
+}
+
+int kvm_arch_remove_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
+{
+ uint8_t t[4];
+ static const uint8_t diag_501[] = {0x83, 0x24, 0x05, 0x01};
+
+ if (cpu_memory_rw_debug(env, bp->pc, t, 4, 0))
+ return -EINVAL;
+ if (memcmp(t, diag_501, 4))
+ return -EINVAL;
+ if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 1))
+ return -EINVAL;
+
+ return 0;
+}
+
+int kvm_arch_pre_run(CPUState *env, struct kvm_run *run)
+{
+ return 0;
+}
+
+int kvm_arch_post_run(CPUState *env, struct kvm_run *run)
+{
+ return 0;
+}
+
+static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
+{
+ struct kvm_s390_interrupt kvmint;
+ int r;
+
+ if (!env->kvm_state)
+ return;
+
+ env->halted = 0;
+ env->exception_index = 0;
+
+ kvmint.type = type;
+ kvmint.parm = parm;
+ kvmint.parm64 = parm64;
+
+ if (vm)
+ r = kvm_vm_ioctl(env->kvm_state, KVM_S390_INTERRUPT, &kvmint);
+ else
+ r = kvm_vcpu_ioctl(env, KVM_S390_INTERRUPT, &kvmint);
+
+ if (r < 0) {
+ fprintf(stderr, "KVM failed to inject interrupt\n");
+ exit(1);
+ }
+}
+
+void kvm_s390_virtio_irq(CPUState *env, int config_change, uint64_t token)
+{
+ _kvm_s390_interrupt(env, KVM_S390_INT_VIRTIO, config_change, token, 1);
+}
+
+static void kvm_s390_interrupt(CPUState *env, int type, uint32_t code)
+{
+ _kvm_s390_interrupt(env, type, code, 0, 0);
+}
+
+static void enter_pgmcheck(CPUState *env, uint16_t code)
+{
+ kvm_s390_interrupt(env, KVM_S390_PROGRAM_INT, code);
+}
+
+static void setcc(CPUState *env, uint64_t cc)
+{
+ env->kvm_run->psw_mask &= ~(3ul << 44);
+ env->kvm_run->psw_mask |= (cc & 3) << 44;
+
+ env->psw.mask &= ~(3ul << 44);
+ env->psw.mask |= (cc & 3) << 44;
+}
+
+static int sclp_service_call(CPUState *env, struct kvm_run *run, uint16_t ipbh0)
+{
+ uint32_t sccb;
+ uint64_t code;
+ int r = 0;
+
+ cpu_synchronize_state(env);
+ sccb = env->regs[ipbh0 & 0xf];
+ code = env->regs[(ipbh0 & 0xf0) >> 4];
+
+ dprintf("sclp(0x%x, 0x%lx)\n", sccb, code);
+
+ if (sccb & ~0x7ffffff8ul) {
+ fprintf(stderr, "KVM: invalid sccb address 0x%x\n", sccb);
+ r = -1;
+ goto out;
+ }
+
+ switch(code) {
+ case 0x00020001:
+ case 0x00120001:
+ stw_phys(sccb + SCP_MEM_CODE, ram_size >> 20);
+ stb_phys(sccb + SCP_INCREMENT, 1);
+ stw_phys(sccb + SCP_RESPONSE_CODE, 0x10);
+ setcc(env, 0);
+
+ _kvm_s390_interrupt(env, KVM_S390_INT_SERVICE, sccb & ~3, 0, 1);
+ break;
+ default:
+ dprintf("KVM: invalid sclp call 0x%x / 0x%lx\n", sccb, code);
+ r = -1;
+ break;
+ }
+
+out:
+ if (r < 0)
+ setcc(env, 3);
+ return 0;
+}
+
+static int handle_priv(CPUState *env, struct kvm_run *run, uint8_t ipa1)
+{
+ int r = 0;
+ uint16_t ipbh0 = (run->s390_sieic.ipb & 0xffff0000) >> 16;
+
+ dprintf("KVM: PRIV: %d\n", ipa1);
+ switch (ipa1) {
+ case PRIV_SCLP_CALL:
+ r = sclp_service_call(env, run, ipbh0);
+ break;
+ default:
+ dprintf("KVM: unknown PRIV: 0x%x\n", ipa1);
+ r = -1;
+ break;
+ }
+
+ return r;
+}
+
+static int handle_hypercall(CPUState *env, struct kvm_run *run)
+{
+ int r;
+
+ cpu_synchronize_state(env);
+ r = s390_virtio_hypercall(env);
+ kvm_arch_put_registers(env);
+
+ return r;
+}
+
+static int handle_diag(CPUState *env, struct kvm_run *run, int ipb_code)
+{
+ int r = 0;
+
+ switch (ipb_code) {
+ case DIAG_KVM_HYPERCALL:
+ r = handle_hypercall(env, run);
+ break;
+ case DIAG_KVM_BREAKPOINT:
+ sleep(10);
+ break;
+ default:
+ dprintf("KVM: unknown DIAG: 0x%x\n", ipb_code);
+ r = -1;
+ break;
+ }
+
+ return r;
+}
+
+static int s390_cpu_restart(CPUState *env)
+{
+ kvm_s390_interrupt(env, KVM_S390_RESTART, 0);
+ env->halted = 0;
+ env->exception_index = 0;
+ qemu_cpu_kick(env);
+ fprintf(stderr, "DONE: SIGP cpu restart: %p\n", env);
+ return 0;
+}
+
+static int s390_store_status(CPUState *env, uint32_t parameter)
+{
+ /* XXX */
+ fprintf(stderr, "XXX SIGP store status\n");
+ return -1;
+}
+
+static int s390_cpu_initial_reset(CPUState *env)
+{
+ /* XXX */
+ fprintf(stderr, "XXX SIGP init\n");
+ return -1;
+}
+
+static int handle_sigp(CPUState *env, struct kvm_run *run, uint8_t ipa1)
+{
+ uint8_t order_code;
+ uint32_t parameter;
+ uint16_t cpu_addr;
+ uint8_t t;
+ int r = -1;
+ CPUState *target_env;
+
+ cpu_synchronize_state(env);
+
+ /* get order code */
+ order_code = run->s390_sieic.ipb >> 28;
+ if (order_code > 0)
+ order_code = env->regs[order_code];
+ order_code += (run->s390_sieic.ipb & 0x0fff0000) >> 16;
+
+ /* get parameters */
+ t = (ipa1 & 0xf0) >> 4;
+ if (!(t % 2))
+ t++;
+
+ parameter = env->regs[t] & 0x7ffffe00;
+ cpu_addr = env->regs[ipa1 & 0x0f];
+
+ target_env = s390_cpu_addr2state(cpu_addr);
+ if (!target_env)
+ goto out;
+
+ switch (order_code) {
+ case SIGP_RESTART:
+ r = s390_cpu_restart(target_env);
+ break;
+ case SIGP_STORE_STATUS_ADDR:
+ r = s390_store_status(target_env, parameter);
+ break;
+ case SIGP_SET_ARCH:
+ /* make the caller panic */
+ return -1;
+ case SIGP_INITIAL_CPU_RESET:
+ r = s390_cpu_initial_reset(target_env);
+ break;
+ default:
+ fprintf(stderr, "KVM: unknown SIGP: 0x%x\n", ipa1);
+ break;
+ }
+
+out:
+ setcc(env, r ? 3 : 0);
+ return 0;
+}
+
+static int handle_instruction(CPUState *env, struct kvm_run *run)
+{
+ unsigned int ipa0 = (run->s390_sieic.ipa & 0xff00);
+ uint8_t ipa1 = run->s390_sieic.ipa & 0x00ff;
+ int ipb_code = (run->s390_sieic.ipb & 0x0fff0000) >> 16;
+ int r = 0;
+
+ dprintf("handle_instruction 0x%x 0x%x\n", run->s390_sieic.ipa, run->s390_sieic.ipb);
+ switch (ipa0) {
+ case IPA0_PRIV:
+ r = handle_priv(env, run, ipa1);
+ break;
+ case IPA0_DIAG:
+ r = handle_diag(env, run, ipb_code);
+ break;
+ case IPA0_SIGP:
+ r = handle_sigp(env, run, ipa1);
+ break;
+ }
+
+ if (r < 0) {
+ enter_pgmcheck(env, 0x0001);
+ }
+ return r;
+}
+
+static int handle_intercept(CPUState *env)
+{
+ struct kvm_run *run = env->kvm_run;
+ int icpt_code = run->s390_sieic.icptcode;
+ int r = 0;
+
+ dprintf("intercept: 0x%x (at 0x%lx)\n", icpt_code, env->kvm_run->psw_addr);
+ switch (icpt_code) {
+ case ICPT_INSTRUCTION:
+ r = handle_instruction(env, run);
+ break;
+ case ICPT_WAITPSW:
+ /* XXX What to do on system shutdown? */
+ env->halted = 1;
+ env->exception_index = EXCP_HLT;
+ break;
+ case ICPT_SOFT_INTERCEPT:
+ fprintf(stderr, "KVM unimplemented icpt SOFT\n");
+ exit(1);
+ break;
+ case ICPT_CPU_STOP:
+ qemu_system_shutdown_request();
+ break;
+ case ICPT_IO:
+ fprintf(stderr, "KVM unimplemented icpt IO\n");
+ exit(1);
+ break;
+ default:
+ fprintf(stderr, "Unknown intercept code: %d\n", icpt_code);
+ exit(1);
+ break;
+ }
+
+ return r;
+}
+
+int kvm_arch_handle_exit(CPUState *env, struct kvm_run *run)
+{
+ int ret = 0;
+
+ switch (run->exit_reason) {
+ case KVM_EXIT_S390_SIEIC:
+ ret = handle_intercept(env);
+ break;
+ case KVM_EXIT_S390_RESET:
+ fprintf(stderr, "RESET not implemented\n");
+ exit(1);
+ break;
+ default:
+ fprintf(stderr, "Unknown KVM exit: %d\n", run->exit_reason);
+ break;
+ }
+
+ return ret;
+}
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-11-26 13:23 ` [Qemu-devel] [PATCH 04/11] Add KVM support for S390x Alexander Graf
@ 2009-11-30 18:18 ` Aurelien Jarno
2009-11-30 22:25 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-11-30 18:18 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
> S390x was one of the first platforms that received support for KVM back in the
> day. Unfortunately until now there hasn't been a qemu implementation that would
> enable users to actually run guests.
>
> So let's include support for KVM S390x in qemu!
Please find the comments below.
> Signed-off-by: Alexander Graf <agraf@suse.de>
> ---
> configure | 4 +-
> target-s390x/kvm.c | 463 ++++++++++++++++++++++++++++++++++++++++++++++++++++
> 2 files changed, 466 insertions(+), 1 deletions(-)
> create mode 100644 target-s390x/kvm.c
>
> diff --git a/configure b/configure
> index b616e1a..cf466ec 100755
> --- a/configure
> +++ b/configure
> @@ -1348,6 +1348,8 @@ EOF
> kvm_cflags="$kvm_cflags -I$kerneldir/arch/x86/include"
> elif test "$cpu" = "ppc" -a -d "$kerneldir/arch/powerpc/include" ; then
> kvm_cflags="$kvm_cflags -I$kerneldir/arch/powerpc/include"
> + elif test "$cpu" = "s390x" -a -d "$kerneldir/arch/s390/include" ; then
> + kvm_cflags="$kvm_cflags -I$kerneldir/arch/s390/include"
> elif test -d "$kerneldir/arch/$cpu/include" ; then
> kvm_cflags="$kvm_cflags -I$kerneldir/arch/$cpu/include"
> fi
> @@ -2360,7 +2362,7 @@ case "$target_arch2" in
> fi
> esac
> case "$target_arch2" in
> - i386|x86_64|ppcemb|ppc|ppc64)
> + i386|x86_64|ppcemb|ppc|ppc64|s390x)
> # Make sure the target and host cpus are compatible
> if test "$kvm" = "yes" -a "$target_softmmu" = "yes" -a \
> \( "$target_arch2" = "$cpu" -o \
> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
> new file mode 100644
> index 0000000..d477664
> --- /dev/null
> +++ b/target-s390x/kvm.c
> @@ -0,0 +1,463 @@
> +/*
> + * QEMU S390x KVM implementation
> + *
> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <sys/types.h>
> +#include <sys/ioctl.h>
> +#include <sys/mman.h>
> +
> +#include <linux/kvm.h>
> +#include <asm/ptrace.h>
> +
> +#include "qemu-common.h"
> +#include "qemu-timer.h"
> +#include "sysemu.h"
> +#include "kvm.h"
> +#include "cpu.h"
> +#include "device_tree.h"
> +
> +/* #define DEBUG_KVM */
> +
> +#ifdef DEBUG_KVM
> +#define dprintf(fmt, ...) \
> + do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
> +#else
> +#define dprintf(fmt, ...) \
> + do { } while (0)
> +#endif
> +
> +#define IPA0_DIAG 0x8300
> +#define IPA0_SIGP 0xae00
> +#define IPA0_PRIV 0xb200
> +
> +#define PRIV_SCLP_CALL 0x20
> +#define DIAG_KVM_HYPERCALL 0x500
> +#define DIAG_KVM_BREAKPOINT 0x501
> +
> +#define SCP_LENGTH 0x00
> +#define SCP_FUNCTION_CODE 0x02
> +#define SCP_CONTROL_MASK 0x03
> +#define SCP_RESPONSE_CODE 0x06
> +#define SCP_MEM_CODE 0x08
> +#define SCP_INCREMENT 0x0a
> +
> +#define ICPT_INSTRUCTION 0x04
> +#define ICPT_WAITPSW 0x1c
> +#define ICPT_SOFT_INTERCEPT 0x24
> +#define ICPT_CPU_STOP 0x28
> +#define ICPT_IO 0x40
> +
> +#define SIGP_RESTART 0x06
> +#define SIGP_INITIAL_CPU_RESET 0x0b
> +#define SIGP_STORE_STATUS_ADDR 0x0e
> +#define SIGP_SET_ARCH 0x12
> +
> +
> +int kvm_arch_init(KVMState *s, int smp_cpus)
> +{
> + return 0;
> +}
> +
> +int kvm_arch_init_vcpu(CPUState *env)
> +{
> + int ret = 0;
> +
> + if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL) < 0)
> + perror("cannot init reset vcpu");
coding style.
> +
> + return ret;
> +}
> +
> +void kvm_arch_reset_vcpu(CPUState *env)
> +{
> +}
Is there really nothing to do? If some code has to be added later, it
may be worth adding a comment.
> +int kvm_arch_put_registers(CPUState *env)
> +{
> + struct kvm_regs regs;
> + int ret;
> + int i;
> +
> + ret = kvm_vcpu_ioctl(env, KVM_GET_REGS, ®s);
> + if (ret < 0)
> + return ret;
coding style.
> + for (i = 0; i < 16; i++)
> + regs.gprs[i] = env->regs[i];
coding style.
> +
> + ret = kvm_vcpu_ioctl(env, KVM_SET_REGS, ®s);
> + if (ret < 0)
> + return ret;
coding style.
> +
> + env->kvm_run->psw_addr = env->psw.addr;
> + env->kvm_run->psw_mask = env->psw.mask;
> +
> + return ret;
> +}
> +
> +int kvm_arch_get_registers(CPUState *env)
> +{
> + uint32_t ret;
> + struct kvm_regs regs;
> + int i;
> +
> + ret = kvm_vcpu_ioctl(env, KVM_GET_REGS, ®s);
> + if (ret < 0)
> + return ret;
coding style.
> +
> + for (i = 0; i < 16; i++)
> + env->regs[i] = regs.gprs[i];
coding style.
> +
> + env->psw.addr = env->kvm_run->psw_addr;
> + env->psw.mask = env->kvm_run->psw_mask;
> +
> + return 0;
> +}
> +
> +int kvm_arch_insert_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
> +{
> + static const uint8_t diag_501[] = {0x83, 0x24, 0x05, 0x01};
> +
> + if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 4, 0) ||
> + cpu_memory_rw_debug(env, bp->pc, (uint8_t *)diag_501, 4, 1))
> + return -EINVAL;
coding style.
> + return 0;
> +}
> +
> +int kvm_arch_remove_sw_breakpoint(CPUState *env, struct kvm_sw_breakpoint *bp)
> +{
> + uint8_t t[4];
> + static const uint8_t diag_501[] = {0x83, 0x24, 0x05, 0x01};
> +
> + if (cpu_memory_rw_debug(env, bp->pc, t, 4, 0))
> + return -EINVAL;
> + if (memcmp(t, diag_501, 4))
> + return -EINVAL;
> + if (cpu_memory_rw_debug(env, bp->pc, (uint8_t *)&bp->saved_insn, 1, 1))
> + return -EINVAL;
coding style.
> +
> + return 0;
> +}
> +
> +int kvm_arch_pre_run(CPUState *env, struct kvm_run *run)
> +{
> + return 0;
> +}
> +
> +int kvm_arch_post_run(CPUState *env, struct kvm_run *run)
> +{
> + return 0;
> +}
> +
> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
> +{
Why such a name starting with an underscore?
> + struct kvm_s390_interrupt kvmint;
> + int r;
> +
> + if (!env->kvm_state)
> + return;
coding style.
> +
> + env->halted = 0;
> + env->exception_index = 0;
> +
> + kvmint.type = type;
> + kvmint.parm = parm;
> + kvmint.parm64 = parm64;
> +
> + if (vm)
> + r = kvm_vm_ioctl(env->kvm_state, KVM_S390_INTERRUPT, &kvmint);
> + else
> + r = kvm_vcpu_ioctl(env, KVM_S390_INTERRUPT, &kvmint);
coding style.
> +
> + if (r < 0) {
> + fprintf(stderr, "KVM failed to inject interrupt\n");
> + exit(1);
> + }
> +}
> +
> +void kvm_s390_virtio_irq(CPUState *env, int config_change, uint64_t token)
> +{
> + _kvm_s390_interrupt(env, KVM_S390_INT_VIRTIO, config_change, token, 1);
> +}
> +
> +static void kvm_s390_interrupt(CPUState *env, int type, uint32_t code)
> +{
> + _kvm_s390_interrupt(env, type, code, 0, 0);
> +}
> +
> +static void enter_pgmcheck(CPUState *env, uint16_t code)
> +{
> + kvm_s390_interrupt(env, KVM_S390_PROGRAM_INT, code);
> +}
> +
> +static void setcc(CPUState *env, uint64_t cc)
> +{
> + env->kvm_run->psw_mask &= ~(3ul << 44);
> + env->kvm_run->psw_mask |= (cc & 3) << 44;
> +
> + env->psw.mask &= ~(3ul << 44);
> + env->psw.mask |= (cc & 3) << 44;
> +}
> +
> +static int sclp_service_call(CPUState *env, struct kvm_run *run, uint16_t ipbh0)
> +{
> + uint32_t sccb;
> + uint64_t code;
> + int r = 0;
> +
> + cpu_synchronize_state(env);
> + sccb = env->regs[ipbh0 & 0xf];
> + code = env->regs[(ipbh0 & 0xf0) >> 4];
> +
> + dprintf("sclp(0x%x, 0x%lx)\n", sccb, code);
> +
> + if (sccb & ~0x7ffffff8ul) {
> + fprintf(stderr, "KVM: invalid sccb address 0x%x\n", sccb);
> + r = -1;
> + goto out;
> + }
> +
> + switch(code) {
> + case 0x00020001:
> + case 0x00120001:
What are those constants?
> + stw_phys(sccb + SCP_MEM_CODE, ram_size >> 20);
> + stb_phys(sccb + SCP_INCREMENT, 1);
> + stw_phys(sccb + SCP_RESPONSE_CODE, 0x10);
> + setcc(env, 0);
> +
> + _kvm_s390_interrupt(env, KVM_S390_INT_SERVICE, sccb & ~3, 0, 1);
> + break;
> + default:
> + dprintf("KVM: invalid sclp call 0x%x / 0x%lx\n", sccb, code);
> + r = -1;
> + break;
> + }
> +
> +out:
> + if (r < 0)
> + setcc(env, 3);
coding style.
> + return 0;
> +}
> +
> +static int handle_priv(CPUState *env, struct kvm_run *run, uint8_t ipa1)
> +{
> + int r = 0;
> + uint16_t ipbh0 = (run->s390_sieic.ipb & 0xffff0000) >> 16;
> +
> + dprintf("KVM: PRIV: %d\n", ipa1);
> + switch (ipa1) {
> + case PRIV_SCLP_CALL:
> + r = sclp_service_call(env, run, ipbh0);
> + break;
> + default:
> + dprintf("KVM: unknown PRIV: 0x%x\n", ipa1);
> + r = -1;
> + break;
> + }
> +
> + return r;
> +}
> +
> +static int handle_hypercall(CPUState *env, struct kvm_run *run)
> +{
> + int r;
> +
> + cpu_synchronize_state(env);
> + r = s390_virtio_hypercall(env);
> + kvm_arch_put_registers(env);
> +
> + return r;
> +}
> +
> +static int handle_diag(CPUState *env, struct kvm_run *run, int ipb_code)
> +{
> + int r = 0;
> +
> + switch (ipb_code) {
> + case DIAG_KVM_HYPERCALL:
> + r = handle_hypercall(env, run);
> + break;
> + case DIAG_KVM_BREAKPOINT:
> + sleep(10);
> + break;
> + default:
> + dprintf("KVM: unknown DIAG: 0x%x\n", ipb_code);
> + r = -1;
> + break;
> + }
> +
> + return r;
> +}
> +
> +static int s390_cpu_restart(CPUState *env)
> +{
> + kvm_s390_interrupt(env, KVM_S390_RESTART, 0);
> + env->halted = 0;
> + env->exception_index = 0;
> + qemu_cpu_kick(env);
> + fprintf(stderr, "DONE: SIGP cpu restart: %p\n", env);
> + return 0;
> +}
> +
> +static int s390_store_status(CPUState *env, uint32_t parameter)
> +{
> + /* XXX */
> + fprintf(stderr, "XXX SIGP store status\n");
> + return -1;
> +}
> +
> +static int s390_cpu_initial_reset(CPUState *env)
> +{
> + /* XXX */
> + fprintf(stderr, "XXX SIGP init\n");
> + return -1;
> +}
> +
> +static int handle_sigp(CPUState *env, struct kvm_run *run, uint8_t ipa1)
> +{
> + uint8_t order_code;
> + uint32_t parameter;
> + uint16_t cpu_addr;
> + uint8_t t;
> + int r = -1;
> + CPUState *target_env;
> +
> + cpu_synchronize_state(env);
> +
> + /* get order code */
> + order_code = run->s390_sieic.ipb >> 28;
> + if (order_code > 0)
> + order_code = env->regs[order_code];
coding style.
> + order_code += (run->s390_sieic.ipb & 0x0fff0000) >> 16;
> +
> + /* get parameters */
> + t = (ipa1 & 0xf0) >> 4;
> + if (!(t % 2))
> + t++;
coding style.
> +
> + parameter = env->regs[t] & 0x7ffffe00;
> + cpu_addr = env->regs[ipa1 & 0x0f];
> +
> + target_env = s390_cpu_addr2state(cpu_addr);
> + if (!target_env)
> + goto out;
coding style.
> +
> + switch (order_code) {
> + case SIGP_RESTART:
> + r = s390_cpu_restart(target_env);
> + break;
> + case SIGP_STORE_STATUS_ADDR:
> + r = s390_store_status(target_env, parameter);
> + break;
> + case SIGP_SET_ARCH:
> + /* make the caller panic */
> + return -1;
> + case SIGP_INITIAL_CPU_RESET:
> + r = s390_cpu_initial_reset(target_env);
> + break;
> + default:
> + fprintf(stderr, "KVM: unknown SIGP: 0x%x\n", ipa1);
> + break;
> + }
> +
> +out:
> + setcc(env, r ? 3 : 0);
> + return 0;
> +}
> +
> +static int handle_instruction(CPUState *env, struct kvm_run *run)
> +{
> + unsigned int ipa0 = (run->s390_sieic.ipa & 0xff00);
> + uint8_t ipa1 = run->s390_sieic.ipa & 0x00ff;
> + int ipb_code = (run->s390_sieic.ipb & 0x0fff0000) >> 16;
> + int r = 0;
> +
> + dprintf("handle_instruction 0x%x 0x%x\n", run->s390_sieic.ipa, run->s390_sieic.ipb);
> + switch (ipa0) {
> + case IPA0_PRIV:
> + r = handle_priv(env, run, ipa1);
> + break;
> + case IPA0_DIAG:
> + r = handle_diag(env, run, ipb_code);
> + break;
> + case IPA0_SIGP:
> + r = handle_sigp(env, run, ipa1);
> + break;
> + }
> +
> + if (r < 0) {
> + enter_pgmcheck(env, 0x0001);
> + }
> + return r;
> +}
> +
> +static int handle_intercept(CPUState *env)
> +{
> + struct kvm_run *run = env->kvm_run;
> + int icpt_code = run->s390_sieic.icptcode;
> + int r = 0;
> +
> + dprintf("intercept: 0x%x (at 0x%lx)\n", icpt_code, env->kvm_run->psw_addr);
> + switch (icpt_code) {
> + case ICPT_INSTRUCTION:
> + r = handle_instruction(env, run);
> + break;
> + case ICPT_WAITPSW:
> + /* XXX What to do on system shutdown? */
> + env->halted = 1;
> + env->exception_index = EXCP_HLT;
> + break;
> + case ICPT_SOFT_INTERCEPT:
> + fprintf(stderr, "KVM unimplemented icpt SOFT\n");
> + exit(1);
> + break;
> + case ICPT_CPU_STOP:
> + qemu_system_shutdown_request();
> + break;
> + case ICPT_IO:
> + fprintf(stderr, "KVM unimplemented icpt IO\n");
> + exit(1);
> + break;
> + default:
> + fprintf(stderr, "Unknown intercept code: %d\n", icpt_code);
> + exit(1);
> + break;
> + }
> +
> + return r;
> +}
> +
> +int kvm_arch_handle_exit(CPUState *env, struct kvm_run *run)
> +{
> + int ret = 0;
> +
> + switch (run->exit_reason) {
> + case KVM_EXIT_S390_SIEIC:
> + ret = handle_intercept(env);
> + break;
> + case KVM_EXIT_S390_RESET:
> + fprintf(stderr, "RESET not implemented\n");
> + exit(1);
> + break;
> + default:
> + fprintf(stderr, "Unknown KVM exit: %d\n", run->exit_reason);
> + break;
> + }
> +
> + return ret;
> +}
> --
> 1.6.0.2
>
>
>
>
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-11-30 18:18 ` Aurelien Jarno
@ 2009-11-30 22:25 ` Alexander Graf
2009-12-02 8:12 ` Aurelien Jarno
0 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-30 22:25 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
>> S390x was one of the first platforms that received support for KVM back in the
>> day. Unfortunately until now there hasn't been a qemu implementation that would
>> enable users to actually run guests.
>>
>> So let's include support for KVM S390x in qemu!
>
> Please find the comments below.
>
>> Signed-off-by: Alexander Graf <agraf@suse.de>
>> ---
>> configure | 4 +-
>> target-s390x/kvm.c | 463 ++++++++++++++++++++++++++++++++++++++++++++++++++++
>> 2 files changed, 466 insertions(+), 1 deletions(-)
>> create mode 100644 target-s390x/kvm.c
>>
>> diff --git a/configure b/configure
>> index b616e1a..cf466ec 100755
>> --- a/configure
>> +++ b/configure
>> @@ -1348,6 +1348,8 @@ EOF
>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/x86/include"
>> elif test "$cpu" = "ppc" -a -d "$kerneldir/arch/powerpc/include" ; then
>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/powerpc/include"
>> + elif test "$cpu" = "s390x" -a -d "$kerneldir/arch/s390/include" ; then
>> + kvm_cflags="$kvm_cflags -I$kerneldir/arch/s390/include"
>> elif test -d "$kerneldir/arch/$cpu/include" ; then
>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/$cpu/include"
>> fi
>> @@ -2360,7 +2362,7 @@ case "$target_arch2" in
>> fi
>> esac
>> case "$target_arch2" in
>> - i386|x86_64|ppcemb|ppc|ppc64)
>> + i386|x86_64|ppcemb|ppc|ppc64|s390x)
>> # Make sure the target and host cpus are compatible
>> if test "$kvm" = "yes" -a "$target_softmmu" = "yes" -a \
>> \( "$target_arch2" = "$cpu" -o \
>> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
>> new file mode 100644
>> index 0000000..d477664
>> --- /dev/null
>> +++ b/target-s390x/kvm.c
>> @@ -0,0 +1,463 @@
>> +/*
>> + * QEMU S390x KVM implementation
>> + *
>> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
>> + *
>> + * This library is free software; you can redistribute it and/or
>> + * modify it under the terms of the GNU Lesser General Public
>> + * License as published by the Free Software Foundation; either
>> + * version 2 of the License, or (at your option) any later version.
>> + *
>> + * This library is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
>> + * Lesser General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU Lesser General Public
>> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
>> + */
>> +
>> +#include <sys/types.h>
>> +#include <sys/ioctl.h>
>> +#include <sys/mman.h>
>> +
>> +#include <linux/kvm.h>
>> +#include <asm/ptrace.h>
>> +
>> +#include "qemu-common.h"
>> +#include "qemu-timer.h"
>> +#include "sysemu.h"
>> +#include "kvm.h"
>> +#include "cpu.h"
>> +#include "device_tree.h"
>> +
>> +/* #define DEBUG_KVM */
>> +
>> +#ifdef DEBUG_KVM
>> +#define dprintf(fmt, ...) \
>> + do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
>> +#else
>> +#define dprintf(fmt, ...) \
>> + do { } while (0)
>> +#endif
>> +
>> +#define IPA0_DIAG 0x8300
>> +#define IPA0_SIGP 0xae00
>> +#define IPA0_PRIV 0xb200
>> +
>> +#define PRIV_SCLP_CALL 0x20
>> +#define DIAG_KVM_HYPERCALL 0x500
>> +#define DIAG_KVM_BREAKPOINT 0x501
>> +
>> +#define SCP_LENGTH 0x00
>> +#define SCP_FUNCTION_CODE 0x02
>> +#define SCP_CONTROL_MASK 0x03
>> +#define SCP_RESPONSE_CODE 0x06
>> +#define SCP_MEM_CODE 0x08
>> +#define SCP_INCREMENT 0x0a
>> +
>> +#define ICPT_INSTRUCTION 0x04
>> +#define ICPT_WAITPSW 0x1c
>> +#define ICPT_SOFT_INTERCEPT 0x24
>> +#define ICPT_CPU_STOP 0x28
>> +#define ICPT_IO 0x40
>> +
>> +#define SIGP_RESTART 0x06
>> +#define SIGP_INITIAL_CPU_RESET 0x0b
>> +#define SIGP_STORE_STATUS_ADDR 0x0e
>> +#define SIGP_SET_ARCH 0x12
>> +
>> +
>> +int kvm_arch_init(KVMState *s, int smp_cpus)
>> +{
>> + return 0;
>> +}
>> +
>> +int kvm_arch_init_vcpu(CPUState *env)
>> +{
>> + int ret = 0;
>> +
>> + if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL) < 0)
>> + perror("cannot init reset vcpu");
>
> coding style.
>
>> +
>> + return ret;
>> +}
>> +
>> +void kvm_arch_reset_vcpu(CPUState *env)
>> +{
>> +}
>
> Is there really nothing to do? If some code has to be added later, it
> may be worth adding a comment.
As you have probably realized by now, all reset code is missing. In fact, we don't even ever call the qemu reset functions to actually do a reset.
>> +
>> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
>> +{
>
> Why such a name starting with an underscore?
Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
>> +static int sclp_service_call(CPUState *env, struct kvm_run *run, uint16_t ipbh0)
>> +{
>> + uint32_t sccb;
>> + uint64_t code;
>> + int r = 0;
>> +
>> + cpu_synchronize_state(env);
>> + sccb = env->regs[ipbh0 & 0xf];
>> + code = env->regs[(ipbh0 & 0xf0) >> 4];
>> +
>> + dprintf("sclp(0x%x, 0x%lx)\n", sccb, code);
>> +
>> + if (sccb & ~0x7ffffff8ul) {
>> + fprintf(stderr, "KVM: invalid sccb address 0x%x\n", sccb);
>> + r = -1;
>> + goto out;
>> + }
>> +
>> + switch(code) {
>> + case 0x00020001:
>> + case 0x00120001:
>
> What are those constants?
If I knew I'd have defined some more verbose constants. I just took them 1:1 from kuli.
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-11-30 22:25 ` Alexander Graf
@ 2009-12-02 8:12 ` Aurelien Jarno
2009-12-02 8:28 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-12-02 8:12 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Mon, Nov 30, 2009 at 11:25:03PM +0100, Alexander Graf wrote:
>
> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>
> > On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
> >> S390x was one of the first platforms that received support for KVM back in the
> >> day. Unfortunately until now there hasn't been a qemu implementation that would
> >> enable users to actually run guests.
> >>
> >> So let's include support for KVM S390x in qemu!
> >
> > Please find the comments below.
> >
> >> Signed-off-by: Alexander Graf <agraf@suse.de>
> >> ---
> >> configure | 4 +-
> >> target-s390x/kvm.c | 463 ++++++++++++++++++++++++++++++++++++++++++++++++++++
> >> 2 files changed, 466 insertions(+), 1 deletions(-)
> >> create mode 100644 target-s390x/kvm.c
> >>
> >> diff --git a/configure b/configure
> >> index b616e1a..cf466ec 100755
> >> --- a/configure
> >> +++ b/configure
> >> @@ -1348,6 +1348,8 @@ EOF
> >> kvm_cflags="$kvm_cflags -I$kerneldir/arch/x86/include"
> >> elif test "$cpu" = "ppc" -a -d "$kerneldir/arch/powerpc/include" ; then
> >> kvm_cflags="$kvm_cflags -I$kerneldir/arch/powerpc/include"
> >> + elif test "$cpu" = "s390x" -a -d "$kerneldir/arch/s390/include" ; then
> >> + kvm_cflags="$kvm_cflags -I$kerneldir/arch/s390/include"
> >> elif test -d "$kerneldir/arch/$cpu/include" ; then
> >> kvm_cflags="$kvm_cflags -I$kerneldir/arch/$cpu/include"
> >> fi
> >> @@ -2360,7 +2362,7 @@ case "$target_arch2" in
> >> fi
> >> esac
> >> case "$target_arch2" in
> >> - i386|x86_64|ppcemb|ppc|ppc64)
> >> + i386|x86_64|ppcemb|ppc|ppc64|s390x)
> >> # Make sure the target and host cpus are compatible
> >> if test "$kvm" = "yes" -a "$target_softmmu" = "yes" -a \
> >> \( "$target_arch2" = "$cpu" -o \
> >> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
> >> new file mode 100644
> >> index 0000000..d477664
> >> --- /dev/null
> >> +++ b/target-s390x/kvm.c
> >> @@ -0,0 +1,463 @@
> >> +/*
> >> + * QEMU S390x KVM implementation
> >> + *
> >> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
> >> + *
> >> + * This library is free software; you can redistribute it and/or
> >> + * modify it under the terms of the GNU Lesser General Public
> >> + * License as published by the Free Software Foundation; either
> >> + * version 2 of the License, or (at your option) any later version.
> >> + *
> >> + * This library is distributed in the hope that it will be useful,
> >> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> >> + * Lesser General Public License for more details.
> >> + *
> >> + * You should have received a copy of the GNU Lesser General Public
> >> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
> >> + */
> >> +
> >> +#include <sys/types.h>
> >> +#include <sys/ioctl.h>
> >> +#include <sys/mman.h>
> >> +
> >> +#include <linux/kvm.h>
> >> +#include <asm/ptrace.h>
> >> +
> >> +#include "qemu-common.h"
> >> +#include "qemu-timer.h"
> >> +#include "sysemu.h"
> >> +#include "kvm.h"
> >> +#include "cpu.h"
> >> +#include "device_tree.h"
> >> +
> >> +/* #define DEBUG_KVM */
> >> +
> >> +#ifdef DEBUG_KVM
> >> +#define dprintf(fmt, ...) \
> >> + do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
> >> +#else
> >> +#define dprintf(fmt, ...) \
> >> + do { } while (0)
> >> +#endif
> >> +
> >> +#define IPA0_DIAG 0x8300
> >> +#define IPA0_SIGP 0xae00
> >> +#define IPA0_PRIV 0xb200
> >> +
> >> +#define PRIV_SCLP_CALL 0x20
> >> +#define DIAG_KVM_HYPERCALL 0x500
> >> +#define DIAG_KVM_BREAKPOINT 0x501
> >> +
> >> +#define SCP_LENGTH 0x00
> >> +#define SCP_FUNCTION_CODE 0x02
> >> +#define SCP_CONTROL_MASK 0x03
> >> +#define SCP_RESPONSE_CODE 0x06
> >> +#define SCP_MEM_CODE 0x08
> >> +#define SCP_INCREMENT 0x0a
> >> +
> >> +#define ICPT_INSTRUCTION 0x04
> >> +#define ICPT_WAITPSW 0x1c
> >> +#define ICPT_SOFT_INTERCEPT 0x24
> >> +#define ICPT_CPU_STOP 0x28
> >> +#define ICPT_IO 0x40
> >> +
> >> +#define SIGP_RESTART 0x06
> >> +#define SIGP_INITIAL_CPU_RESET 0x0b
> >> +#define SIGP_STORE_STATUS_ADDR 0x0e
> >> +#define SIGP_SET_ARCH 0x12
> >> +
> >> +
> >> +int kvm_arch_init(KVMState *s, int smp_cpus)
> >> +{
> >> + return 0;
> >> +}
> >> +
> >> +int kvm_arch_init_vcpu(CPUState *env)
> >> +{
> >> + int ret = 0;
> >> +
> >> + if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL) < 0)
> >> + perror("cannot init reset vcpu");
> >
> > coding style.
> >
> >> +
> >> + return ret;
> >> +}
> >> +
> >> +void kvm_arch_reset_vcpu(CPUState *env)
> >> +{
> >> +}
> >
> > Is there really nothing to do? If some code has to be added later, it
> > may be worth adding a comment.
>
> As you have probably realized by now, all reset code is missing. In fact, we don't even ever call the qemu reset functions to actually do a reset.
Can you make it clean with an abort?
> >> +
> >> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
> >> +{
> >
> > Why such a name starting with an underscore?
>
> Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
I don't think there is any convention, but I know malc always complains
about not introducing names starting with an underscore.
> >> +static int sclp_service_call(CPUState *env, struct kvm_run *run, uint16_t ipbh0)
> >> +{
> >> + uint32_t sccb;
> >> + uint64_t code;
> >> + int r = 0;
> >> +
> >> + cpu_synchronize_state(env);
> >> + sccb = env->regs[ipbh0 & 0xf];
> >> + code = env->regs[(ipbh0 & 0xf0) >> 4];
> >> +
> >> + dprintf("sclp(0x%x, 0x%lx)\n", sccb, code);
> >> +
> >> + if (sccb & ~0x7ffffff8ul) {
> >> + fprintf(stderr, "KVM: invalid sccb address 0x%x\n", sccb);
> >> + r = -1;
> >> + goto out;
> >> + }
> >> +
> >> + switch(code) {
> >> + case 0x00020001:
> >> + case 0x00120001:
> >
> > What are those constants?
>
> If I knew I'd have defined some more verbose constants. I just took them 1:1 from kuli.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-12-02 8:12 ` Aurelien Jarno
@ 2009-12-02 8:28 ` Alexander Graf
2009-12-02 8:42 ` malc
0 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-12-02 8:28 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 02.12.2009, at 09:12, Aurelien Jarno wrote:
> On Mon, Nov 30, 2009 at 11:25:03PM +0100, Alexander Graf wrote:
>>
>> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>>
>>> On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
>>>> S390x was one of the first platforms that received support for KVM back in the
>>>> day. Unfortunately until now there hasn't been a qemu implementation that would
>>>> enable users to actually run guests.
>>>>
>>>> So let's include support for KVM S390x in qemu!
>>>
>>> Please find the comments below.
>>>
>>>> Signed-off-by: Alexander Graf <agraf@suse.de>
>>>> ---
>>>> configure | 4 +-
>>>> target-s390x/kvm.c | 463 ++++++++++++++++++++++++++++++++++++++++++++++++++++
>>>> 2 files changed, 466 insertions(+), 1 deletions(-)
>>>> create mode 100644 target-s390x/kvm.c
>>>>
>>>> diff --git a/configure b/configure
>>>> index b616e1a..cf466ec 100755
>>>> --- a/configure
>>>> +++ b/configure
>>>> @@ -1348,6 +1348,8 @@ EOF
>>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/x86/include"
>>>> elif test "$cpu" = "ppc" -a -d "$kerneldir/arch/powerpc/include" ; then
>>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/powerpc/include"
>>>> + elif test "$cpu" = "s390x" -a -d "$kerneldir/arch/s390/include" ; then
>>>> + kvm_cflags="$kvm_cflags -I$kerneldir/arch/s390/include"
>>>> elif test -d "$kerneldir/arch/$cpu/include" ; then
>>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/$cpu/include"
>>>> fi
>>>> @@ -2360,7 +2362,7 @@ case "$target_arch2" in
>>>> fi
>>>> esac
>>>> case "$target_arch2" in
>>>> - i386|x86_64|ppcemb|ppc|ppc64)
>>>> + i386|x86_64|ppcemb|ppc|ppc64|s390x)
>>>> # Make sure the target and host cpus are compatible
>>>> if test "$kvm" = "yes" -a "$target_softmmu" = "yes" -a \
>>>> \( "$target_arch2" = "$cpu" -o \
>>>> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
>>>> new file mode 100644
>>>> index 0000000..d477664
>>>> --- /dev/null
>>>> +++ b/target-s390x/kvm.c
>>>> @@ -0,0 +1,463 @@
>>>> +/*
>>>> + * QEMU S390x KVM implementation
>>>> + *
>>>> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
>>>> + *
>>>> + * This library is free software; you can redistribute it and/or
>>>> + * modify it under the terms of the GNU Lesser General Public
>>>> + * License as published by the Free Software Foundation; either
>>>> + * version 2 of the License, or (at your option) any later version.
>>>> + *
>>>> + * This library is distributed in the hope that it will be useful,
>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
>>>> + * Lesser General Public License for more details.
>>>> + *
>>>> + * You should have received a copy of the GNU Lesser General Public
>>>> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
>>>> + */
>>>> +
>>>> +#include <sys/types.h>
>>>> +#include <sys/ioctl.h>
>>>> +#include <sys/mman.h>
>>>> +
>>>> +#include <linux/kvm.h>
>>>> +#include <asm/ptrace.h>
>>>> +
>>>> +#include "qemu-common.h"
>>>> +#include "qemu-timer.h"
>>>> +#include "sysemu.h"
>>>> +#include "kvm.h"
>>>> +#include "cpu.h"
>>>> +#include "device_tree.h"
>>>> +
>>>> +/* #define DEBUG_KVM */
>>>> +
>>>> +#ifdef DEBUG_KVM
>>>> +#define dprintf(fmt, ...) \
>>>> + do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
>>>> +#else
>>>> +#define dprintf(fmt, ...) \
>>>> + do { } while (0)
>>>> +#endif
>>>> +
>>>> +#define IPA0_DIAG 0x8300
>>>> +#define IPA0_SIGP 0xae00
>>>> +#define IPA0_PRIV 0xb200
>>>> +
>>>> +#define PRIV_SCLP_CALL 0x20
>>>> +#define DIAG_KVM_HYPERCALL 0x500
>>>> +#define DIAG_KVM_BREAKPOINT 0x501
>>>> +
>>>> +#define SCP_LENGTH 0x00
>>>> +#define SCP_FUNCTION_CODE 0x02
>>>> +#define SCP_CONTROL_MASK 0x03
>>>> +#define SCP_RESPONSE_CODE 0x06
>>>> +#define SCP_MEM_CODE 0x08
>>>> +#define SCP_INCREMENT 0x0a
>>>> +
>>>> +#define ICPT_INSTRUCTION 0x04
>>>> +#define ICPT_WAITPSW 0x1c
>>>> +#define ICPT_SOFT_INTERCEPT 0x24
>>>> +#define ICPT_CPU_STOP 0x28
>>>> +#define ICPT_IO 0x40
>>>> +
>>>> +#define SIGP_RESTART 0x06
>>>> +#define SIGP_INITIAL_CPU_RESET 0x0b
>>>> +#define SIGP_STORE_STATUS_ADDR 0x0e
>>>> +#define SIGP_SET_ARCH 0x12
>>>> +
>>>> +
>>>> +int kvm_arch_init(KVMState *s, int smp_cpus)
>>>> +{
>>>> + return 0;
>>>> +}
>>>> +
>>>> +int kvm_arch_init_vcpu(CPUState *env)
>>>> +{
>>>> + int ret = 0;
>>>> +
>>>> + if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL) < 0)
>>>> + perror("cannot init reset vcpu");
>>>
>>> coding style.
>>>
>>>> +
>>>> + return ret;
>>>> +}
>>>> +
>>>> +void kvm_arch_reset_vcpu(CPUState *env)
>>>> +{
>>>> +}
>>>
>>> Is there really nothing to do? If some code has to be added later, it
>>> may be worth adding a comment.
>>
>> As you have probably realized by now, all reset code is missing. In fact, we don't even ever call the qemu reset functions to actually do a reset.
>
> Can you make it clean with an abort?
>
>>>> +
>>>> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
>>>> +{
>>>
>>> Why such a name starting with an underscore?
>>
>> Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
>
> I don't think there is any convention, but I know malc always complains
> about not introducing names starting with an underscore.
Hm - I just wanted to clearly show that this is an internal API, nobody should really have to call directly. But I'm open for other naming suggestions.
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-12-02 8:28 ` Alexander Graf
@ 2009-12-02 8:42 ` malc
2009-12-02 8:47 ` Alexander Graf
2009-12-02 9:36 ` Markus Armbruster
0 siblings, 2 replies; 49+ messages in thread
From: malc @ 2009-12-02 8:42 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel, Aurelien Jarno
On Wed, 2 Dec 2009, Alexander Graf wrote:
>
> On 02.12.2009, at 09:12, Aurelien Jarno wrote:
>
> > On Mon, Nov 30, 2009 at 11:25:03PM +0100, Alexander Graf wrote:
> >>
> >> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> >>
> >>> On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
> >>>> S390x was one of the first platforms that received support for KVM back in the
> >>>> day. Unfortunately until now there hasn't been a qemu implementation that would
> >>>> enable users to actually run guests.
> >>>>
> >>>> So let's include support for KVM S390x in qemu!
> >>>
> >>> Please find the comments below.
> >>>
> >>>> Signed-off-by: Alexander Graf <agraf@suse.de>
> >>>> ---
> >>>> configure | 4 +-
> >>>> target-s390x/kvm.c | 463 ++++++++++++++++++++++++++++++++++++++++++++++++++++
> >>>> 2 files changed, 466 insertions(+), 1 deletions(-)
> >>>> create mode 100644 target-s390x/kvm.c
> >>>>
> >>>> diff --git a/configure b/configure
> >>>> index b616e1a..cf466ec 100755
> >>>> --- a/configure
> >>>> +++ b/configure
> >>>> @@ -1348,6 +1348,8 @@ EOF
> >>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/x86/include"
> >>>> elif test "$cpu" = "ppc" -a -d "$kerneldir/arch/powerpc/include" ; then
> >>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/powerpc/include"
> >>>> + elif test "$cpu" = "s390x" -a -d "$kerneldir/arch/s390/include" ; then
> >>>> + kvm_cflags="$kvm_cflags -I$kerneldir/arch/s390/include"
> >>>> elif test -d "$kerneldir/arch/$cpu/include" ; then
> >>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/$cpu/include"
> >>>> fi
> >>>> @@ -2360,7 +2362,7 @@ case "$target_arch2" in
> >>>> fi
> >>>> esac
> >>>> case "$target_arch2" in
> >>>> - i386|x86_64|ppcemb|ppc|ppc64)
> >>>> + i386|x86_64|ppcemb|ppc|ppc64|s390x)
> >>>> # Make sure the target and host cpus are compatible
> >>>> if test "$kvm" = "yes" -a "$target_softmmu" = "yes" -a \
> >>>> \( "$target_arch2" = "$cpu" -o \
> >>>> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
> >>>> new file mode 100644
> >>>> index 0000000..d477664
> >>>> --- /dev/null
> >>>> +++ b/target-s390x/kvm.c
> >>>> @@ -0,0 +1,463 @@
> >>>> +/*
> >>>> + * QEMU S390x KVM implementation
> >>>> + *
> >>>> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
> >>>> + *
> >>>> + * This library is free software; you can redistribute it and/or
> >>>> + * modify it under the terms of the GNU Lesser General Public
> >>>> + * License as published by the Free Software Foundation; either
> >>>> + * version 2 of the License, or (at your option) any later version.
> >>>> + *
> >>>> + * This library is distributed in the hope that it will be useful,
> >>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> >>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> >>>> + * Lesser General Public License for more details.
> >>>> + *
> >>>> + * You should have received a copy of the GNU Lesser General Public
> >>>> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
> >>>> + */
> >>>> +
> >>>> +#include <sys/types.h>
> >>>> +#include <sys/ioctl.h>
> >>>> +#include <sys/mman.h>
> >>>> +
> >>>> +#include <linux/kvm.h>
> >>>> +#include <asm/ptrace.h>
> >>>> +
> >>>> +#include "qemu-common.h"
> >>>> +#include "qemu-timer.h"
> >>>> +#include "sysemu.h"
> >>>> +#include "kvm.h"
> >>>> +#include "cpu.h"
> >>>> +#include "device_tree.h"
> >>>> +
> >>>> +/* #define DEBUG_KVM */
> >>>> +
> >>>> +#ifdef DEBUG_KVM
> >>>> +#define dprintf(fmt, ...) \
> >>>> + do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
> >>>> +#else
> >>>> +#define dprintf(fmt, ...) \
> >>>> + do { } while (0)
> >>>> +#endif
> >>>> +
> >>>> +#define IPA0_DIAG 0x8300
> >>>> +#define IPA0_SIGP 0xae00
> >>>> +#define IPA0_PRIV 0xb200
> >>>> +
> >>>> +#define PRIV_SCLP_CALL 0x20
> >>>> +#define DIAG_KVM_HYPERCALL 0x500
> >>>> +#define DIAG_KVM_BREAKPOINT 0x501
> >>>> +
> >>>> +#define SCP_LENGTH 0x00
> >>>> +#define SCP_FUNCTION_CODE 0x02
> >>>> +#define SCP_CONTROL_MASK 0x03
> >>>> +#define SCP_RESPONSE_CODE 0x06
> >>>> +#define SCP_MEM_CODE 0x08
> >>>> +#define SCP_INCREMENT 0x0a
> >>>> +
> >>>> +#define ICPT_INSTRUCTION 0x04
> >>>> +#define ICPT_WAITPSW 0x1c
> >>>> +#define ICPT_SOFT_INTERCEPT 0x24
> >>>> +#define ICPT_CPU_STOP 0x28
> >>>> +#define ICPT_IO 0x40
> >>>> +
> >>>> +#define SIGP_RESTART 0x06
> >>>> +#define SIGP_INITIAL_CPU_RESET 0x0b
> >>>> +#define SIGP_STORE_STATUS_ADDR 0x0e
> >>>> +#define SIGP_SET_ARCH 0x12
> >>>> +
> >>>> +
> >>>> +int kvm_arch_init(KVMState *s, int smp_cpus)
> >>>> +{
> >>>> + return 0;
> >>>> +}
> >>>> +
> >>>> +int kvm_arch_init_vcpu(CPUState *env)
> >>>> +{
> >>>> + int ret = 0;
> >>>> +
> >>>> + if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL) < 0)
> >>>> + perror("cannot init reset vcpu");
> >>>
> >>> coding style.
> >>>
> >>>> +
> >>>> + return ret;
> >>>> +}
> >>>> +
> >>>> +void kvm_arch_reset_vcpu(CPUState *env)
> >>>> +{
> >>>> +}
> >>>
> >>> Is there really nothing to do? If some code has to be added later, it
> >>> may be worth adding a comment.
> >>
> >> As you have probably realized by now, all reset code is missing. In fact, we don't even ever call the qemu reset functions to actually do a reset.
> >
> > Can you make it clean with an abort?
> >
> >>>> +
> >>>> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
> >>>> +{
> >>>
> >>> Why such a name starting with an underscore?
> >>
> >> Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
> >
> > I don't think there is any convention, but I know malc always complains
> > about not introducing names starting with an underscore.
Yeah he does.
>
> Hm - I just wanted to clearly show that this is an internal API, nobody
> should really have to call directly. But I'm open for other naming
> suggestions.
Thing is, in 7.1.3#1 standard says (after explicitly reserving __ _[A-Z]
for any use):
-- All identifiers that begin with an underscore are
always reserved for use as identifiers with file scope
in both the ordinary and tag name spaces.
And i could never really understand (or recall/comprehend when asked
and being given an answer) what this entails. (Anyone?)
So i would go with something imaginative like internal_do_not_use_kvm*,
but that's just me. You can go wild here, leading underscore doesn't look
attractive though.
--
mailto:av1474@comtv.ru
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-12-02 8:42 ` malc
@ 2009-12-02 8:47 ` Alexander Graf
2009-12-02 8:59 ` malc
2009-12-02 9:36 ` Markus Armbruster
1 sibling, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-12-02 8:47 UTC (permalink / raw)
To: malc; +Cc: Carsten Otte, qemu-devel, Aurelien Jarno
[-- Attachment #1: Type: text/plain, Size: 7381 bytes --]
On 02.12.2009, at 09:42, malc wrote:
> On Wed, 2 Dec 2009, Alexander Graf wrote:
>
>>
>> On 02.12.2009, at 09:12, Aurelien Jarno wrote:
>>
>>> On Mon, Nov 30, 2009 at 11:25:03PM +0100, Alexander Graf wrote:
>>>>
>>>> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>>>>
>>>>> On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
>>>>>> S390x was one of the first platforms that received support for KVM back in the
>>>>>> day. Unfortunately until now there hasn't been a qemu implementation that would
>>>>>> enable users to actually run guests.
>>>>>>
>>>>>> So let's include support for KVM S390x in qemu!
>>>>>
>>>>> Please find the comments below.
>>>>>
>>>>>> Signed-off-by: Alexander Graf <agraf@suse.de>
>>>>>> ---
>>>>>> configure | 4 +-
>>>>>> target-s390x/kvm.c | 463 ++++++++++++++++++++++++++++++++++++++++++++++++++++
>>>>>> 2 files changed, 466 insertions(+), 1 deletions(-)
>>>>>> create mode 100644 target-s390x/kvm.c
>>>>>>
>>>>>> diff --git a/configure b/configure
>>>>>> index b616e1a..cf466ec 100755
>>>>>> --- a/configure
>>>>>> +++ b/configure
>>>>>> @@ -1348,6 +1348,8 @@ EOF
>>>>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/x86/include"
>>>>>> elif test "$cpu" = "ppc" -a -d "$kerneldir/arch/powerpc/include" ; then
>>>>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/powerpc/include"
>>>>>> + elif test "$cpu" = "s390x" -a -d "$kerneldir/arch/s390/include" ; then
>>>>>> + kvm_cflags="$kvm_cflags -I$kerneldir/arch/s390/include"
>>>>>> elif test -d "$kerneldir/arch/$cpu/include" ; then
>>>>>> kvm_cflags="$kvm_cflags -I$kerneldir/arch/$cpu/include"
>>>>>> fi
>>>>>> @@ -2360,7 +2362,7 @@ case "$target_arch2" in
>>>>>> fi
>>>>>> esac
>>>>>> case "$target_arch2" in
>>>>>> - i386|x86_64|ppcemb|ppc|ppc64)
>>>>>> + i386|x86_64|ppcemb|ppc|ppc64|s390x)
>>>>>> # Make sure the target and host cpus are compatible
>>>>>> if test "$kvm" = "yes" -a "$target_softmmu" = "yes" -a \
>>>>>> \( "$target_arch2" = "$cpu" -o \
>>>>>> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
>>>>>> new file mode 100644
>>>>>> index 0000000..d477664
>>>>>> --- /dev/null
>>>>>> +++ b/target-s390x/kvm.c
>>>>>> @@ -0,0 +1,463 @@
>>>>>> +/*
>>>>>> + * QEMU S390x KVM implementation
>>>>>> + *
>>>>>> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
>>>>>> + *
>>>>>> + * This library is free software; you can redistribute it and/or
>>>>>> + * modify it under the terms of the GNU Lesser General Public
>>>>>> + * License as published by the Free Software Foundation; either
>>>>>> + * version 2 of the License, or (at your option) any later version.
>>>>>> + *
>>>>>> + * This library is distributed in the hope that it will be useful,
>>>>>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>>>>>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
>>>>>> + * Lesser General Public License for more details.
>>>>>> + *
>>>>>> + * You should have received a copy of the GNU Lesser General Public
>>>>>> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
>>>>>> + */
>>>>>> +
>>>>>> +#include <sys/types.h>
>>>>>> +#include <sys/ioctl.h>
>>>>>> +#include <sys/mman.h>
>>>>>> +
>>>>>> +#include <linux/kvm.h>
>>>>>> +#include <asm/ptrace.h>
>>>>>> +
>>>>>> +#include "qemu-common.h"
>>>>>> +#include "qemu-timer.h"
>>>>>> +#include "sysemu.h"
>>>>>> +#include "kvm.h"
>>>>>> +#include "cpu.h"
>>>>>> +#include "device_tree.h"
>>>>>> +
>>>>>> +/* #define DEBUG_KVM */
>>>>>> +
>>>>>> +#ifdef DEBUG_KVM
>>>>>> +#define dprintf(fmt, ...) \
>>>>>> + do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
>>>>>> +#else
>>>>>> +#define dprintf(fmt, ...) \
>>>>>> + do { } while (0)
>>>>>> +#endif
>>>>>> +
>>>>>> +#define IPA0_DIAG 0x8300
>>>>>> +#define IPA0_SIGP 0xae00
>>>>>> +#define IPA0_PRIV 0xb200
>>>>>> +
>>>>>> +#define PRIV_SCLP_CALL 0x20
>>>>>> +#define DIAG_KVM_HYPERCALL 0x500
>>>>>> +#define DIAG_KVM_BREAKPOINT 0x501
>>>>>> +
>>>>>> +#define SCP_LENGTH 0x00
>>>>>> +#define SCP_FUNCTION_CODE 0x02
>>>>>> +#define SCP_CONTROL_MASK 0x03
>>>>>> +#define SCP_RESPONSE_CODE 0x06
>>>>>> +#define SCP_MEM_CODE 0x08
>>>>>> +#define SCP_INCREMENT 0x0a
>>>>>> +
>>>>>> +#define ICPT_INSTRUCTION 0x04
>>>>>> +#define ICPT_WAITPSW 0x1c
>>>>>> +#define ICPT_SOFT_INTERCEPT 0x24
>>>>>> +#define ICPT_CPU_STOP 0x28
>>>>>> +#define ICPT_IO 0x40
>>>>>> +
>>>>>> +#define SIGP_RESTART 0x06
>>>>>> +#define SIGP_INITIAL_CPU_RESET 0x0b
>>>>>> +#define SIGP_STORE_STATUS_ADDR 0x0e
>>>>>> +#define SIGP_SET_ARCH 0x12
>>>>>> +
>>>>>> +
>>>>>> +int kvm_arch_init(KVMState *s, int smp_cpus)
>>>>>> +{
>>>>>> + return 0;
>>>>>> +}
>>>>>> +
>>>>>> +int kvm_arch_init_vcpu(CPUState *env)
>>>>>> +{
>>>>>> + int ret = 0;
>>>>>> +
>>>>>> + if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL) < 0)
>>>>>> + perror("cannot init reset vcpu");
>>>>>
>>>>> coding style.
>>>>>
>>>>>> +
>>>>>> + return ret;
>>>>>> +}
>>>>>> +
>>>>>> +void kvm_arch_reset_vcpu(CPUState *env)
>>>>>> +{
>>>>>> +}
>>>>>
>>>>> Is there really nothing to do? If some code has to be added later, it
>>>>> may be worth adding a comment.
>>>>
>>>> As you have probably realized by now, all reset code is missing. In fact, we don't even ever call the qemu reset functions to actually do a reset.
>>>
>>> Can you make it clean with an abort?
>>>
>>>>>> +
>>>>>> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
>>>>>> +{
>>>>>
>>>>> Why such a name starting with an underscore?
>>>>
>>>> Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
>>>
>>> I don't think there is any convention, but I know malc always complains
>>> about not introducing names starting with an underscore.
>
> Yeah he does.
>
>>
>> Hm - I just wanted to clearly show that this is an internal API, nobody
>> should really have to call directly. But I'm open for other naming
>> suggestions.
>
> Thing is, in 7.1.3#1 standard says (after explicitly reserving __ _[A-Z]
> for any use):
> -- All identifiers that begin with an underscore are
> always reserved for use as identifiers with file scope
> in both the ordinary and tag name spaces.
>
> And i could never really understand (or recall/comprehend when asked
> and being given an answer) what this entails. (Anyone?)
I don't get the second part, but the first one clearly says "If you use a function beginning underscore, only use it within the file you're declaring it at".
> So i would go with something imaginative like internal_do_not_use_kvm*,
> but that's just me. You can go wild here, leading underscore doesn't look
> attractive though.
Well I could have gone with kvm_s390_interrupt_generic or kvm_s390_interrupt_internal. But I wanted to have a function name that doesn't exceed the 80 character limit all by itself ;-).
Alex
[-- Attachment #2: Type: text/html, Size: 36770 bytes --]
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-12-02 8:47 ` Alexander Graf
@ 2009-12-02 8:59 ` malc
0 siblings, 0 replies; 49+ messages in thread
From: malc @ 2009-12-02 8:59 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel, Aurelien Jarno
On Wed, 2 Dec 2009, Alexander Graf wrote:
>
> On 02.12.2009, at 09:42, malc wrote:
>
> > On Wed, 2 Dec 2009, Alexander Graf wrote:
[..snip..]
> >>>>>
> >>>>> Why such a name starting with an underscore?
> >>>>
> >>>> Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
> >>>
> >>> I don't think there is any convention, but I know malc always complains
> >>> about not introducing names starting with an underscore.
> >
> > Yeah he does.
> >
> >>
> >> Hm - I just wanted to clearly show that this is an internal API, nobody
> >> should really have to call directly. But I'm open for other naming
> >> suggestions.
> >
> > Thing is, in 7.1.3#1 standard says (after explicitly reserving __ _[A-Z]
> > for any use):
> > -- All identifiers that begin with an underscore are
> > always reserved for use as identifiers with file scope
> > in both the ordinary and tag name spaces.
> >
> > And i could never really understand (or recall/comprehend when asked
> > and being given an answer) what this entails. (Anyone?)
>
> I don't get the second part, but the first one clearly says "If you
> use a function beginning underscore, only use it within the file
> you're declaring it at".
6.2.3 describes name spaces.
FWIW here's my interpretation: those names are reserved by implementation
for use in oridnary and tag name spaces, so keep out, feel free to use them
in other name spaces.
Perhaps Mans Rullgard, if he still reads this that is, can enlighten
us. Unfortuantelly i do not have write access to Usenet anymore so
can not ask this on comp.std.c on my own.
>
> > So i would go with something imaginative like internal_do_not_use_kvm*,
> > but that's just me. You can go wild here, leading underscore doesn't look
> > attractive though.
>
> Well I could have gone with kvm_s390_interrupt_generic or
> kvm_s390_interrupt_internal. But I wanted to have a function name
> that doesn't exceed the 80 character limit all by itself ;-).
Oh c'mon those are less than 30 chars :)
--
mailto:av1474@comtv.ru
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-12-02 8:42 ` malc
2009-12-02 8:47 ` Alexander Graf
@ 2009-12-02 9:36 ` Markus Armbruster
2009-12-02 9:48 ` malc
1 sibling, 1 reply; 49+ messages in thread
From: Markus Armbruster @ 2009-12-02 9:36 UTC (permalink / raw)
To: malc; +Cc: Carsten Otte, Alexander Graf, Aurelien Jarno, qemu-devel
malc <av1474@comtv.ru> writes:
> On Wed, 2 Dec 2009, Alexander Graf wrote:
>
>>
>> On 02.12.2009, at 09:12, Aurelien Jarno wrote:
>>
>> > On Mon, Nov 30, 2009 at 11:25:03PM +0100, Alexander Graf wrote:
>> >>
>> >> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>> >>
>> >>> On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
[...]
>> >>>> +
>> >>>> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
>> >>>> +{
>> >>>
>> >>> Why such a name starting with an underscore?
>> >>
>> >> Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
>> >
>> > I don't think there is any convention, but I know malc always complains
>> > about not introducing names starting with an underscore.
>
> Yeah he does.
>
>>
>> Hm - I just wanted to clearly show that this is an internal API, nobody
>> should really have to call directly. But I'm open for other naming
>> suggestions.
>
> Thing is, in 7.1.3#1 standard says (after explicitly reserving __ _[A-Z]
> for any use):
> -- All identifiers that begin with an underscore are
> always reserved for use as identifiers with file scope
> in both the ordinary and tag name spaces.
>
> And i could never really understand (or recall/comprehend when asked
> and being given an answer) what this entails. (Anyone?)
Later in 7.1.3:
If the program declares or defines an identifier in a context in
which it is reserved (other than as allowed by 7.1.4), or defines a
reserved identifier as a macro name, the behavior is undefined.
This gives implementations of the standard (compiler + libc) license to
use reserved identifiers for their own purposes. If they clash with the
user's identifiers, and things break, the user gets to keep the pieces.
Now, it's quite unlikely that _kvm_s390_interrupt() clashes with
anything in practice. It does, however, set a bad example.
> So i would go with something imaginative like internal_do_not_use_kvm*,
> but that's just me. You can go wild here, leading underscore doesn't look
> attractive though.
Why not kvm_s390_interrupt_internal(), or even kvm_s390_interrupt_()?
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 04/11] Add KVM support for S390x
2009-12-02 9:36 ` Markus Armbruster
@ 2009-12-02 9:48 ` malc
0 siblings, 0 replies; 49+ messages in thread
From: malc @ 2009-12-02 9:48 UTC (permalink / raw)
To: Markus Armbruster
Cc: Carsten Otte, Alexander Graf, Aurelien Jarno, qemu-devel
On Wed, 2 Dec 2009, Markus Armbruster wrote:
> malc <av1474@comtv.ru> writes:
>
> > On Wed, 2 Dec 2009, Alexander Graf wrote:
> >
> >>
> >> On 02.12.2009, at 09:12, Aurelien Jarno wrote:
> >>
> >> > On Mon, Nov 30, 2009 at 11:25:03PM +0100, Alexander Graf wrote:
> >> >>
> >> >> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> >> >>
> >> >>> On Thu, Nov 26, 2009 at 02:23:13PM +0100, Alexander Graf wrote:
> [...]
> >> >>>> +
> >> >>>> +static void _kvm_s390_interrupt(CPUState *env, int type, uint32_t parm, uint64_t parm64, int vm)
> >> >>>> +{
> >> >>>
> >> >>> Why such a name starting with an underscore?
> >> >>
> >> >> Because that's the internal function that gets used by the exported, properly named ones. Are there any conventions on how to declare private functions?
> >> >
> >> > I don't think there is any convention, but I know malc always complains
> >> > about not introducing names starting with an underscore.
> >
> > Yeah he does.
> >
> >>
> >> Hm - I just wanted to clearly show that this is an internal API, nobody
> >> should really have to call directly. But I'm open for other naming
> >> suggestions.
> >
> > Thing is, in 7.1.3#1 standard says (after explicitly reserving __ _[A-Z]
> > for any use):
> > -- All identifiers that begin with an underscore are
> > always reserved for use as identifiers with file scope
> > in both the ordinary and tag name spaces.
> >
> > And i could never really understand (or recall/comprehend when asked
> > and being given an answer) what this entails. (Anyone?)
>
> Later in 7.1.3:
>
> If the program declares or defines an identifier in a context in
> which it is reserved (other than as allowed by 7.1.4), or defines a
> reserved identifier as a macro name, the behavior is undefined.
>
> This gives implementations of the standard (compiler + libc) license to
> use reserved identifiers for their own purposes. If they clash with the
> user's identifiers, and things break, the user gets to keep the pieces.
Yes, that's how i would interpret it too (as stated in another message)
>
> Now, it's quite unlikely that _kvm_s390_interrupt() clashes with
> anything in practice. It does, however, set a bad example.
Indeed.
>
> > So i would go with something imaginative like internal_do_not_use_kvm*,
> > but that's just me. You can go wild here, leading underscore doesn't look
> > attractive though.
>
> Why not kvm_s390_interrupt_internal(), or even kvm_s390_interrupt_()?
>
--
mailto:av1474@comtv.ru
^ permalink raw reply [flat|nested] 49+ messages in thread
* [Qemu-devel] [PATCH 05/11] Allocate physical memory in low virtual address space
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (3 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 04/11] Add KVM support for S390x Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-30 18:18 ` Aurelien Jarno
2009-11-26 13:23 ` [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation Alexander Graf
` (6 subsequent siblings)
11 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
KVM on S390x requires the virtual address space of the guest's RAM to be
within the first 256GB.
The general direction I'd like to see KVM on S390 move is that this requirement
is losened, but for now that's what we're stuck with.
So let's just hack up qemu_ram_alloc until KVM behaves nicely :-).
Signed-off-by: Alexander Graf <agraf@suse.de>
---
exec.c | 6 ++++++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/exec.c b/exec.c
index 076d26b..59150d0 100644
--- a/exec.c
+++ b/exec.c
@@ -2411,7 +2411,13 @@ ram_addr_t qemu_ram_alloc(ram_addr_t size)
size = TARGET_PAGE_ALIGN(size);
new_block = qemu_malloc(sizeof(*new_block));
+#if defined(TARGET_S390) && defined(CONFIG_KVM)
+ /* XXX S390 KVM requires the topmost vma of the RAM to be < 256GB */
+ new_block->host = mmap(0x1000000, size, PROT_EXEC|PROT_READ|PROT_WRITE,
+ MAP_SHARED | MAP_ANONYMOUS, -1, 0);
+#else
new_block->host = qemu_vmalloc(size);
+#endif
#ifdef MADV_MERGEABLE
madvise(new_block->host, size, MADV_MERGEABLE);
#endif
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 05/11] Allocate physical memory in low virtual address space
2009-11-26 13:23 ` [Qemu-devel] [PATCH 05/11] Allocate physical memory in low virtual address space Alexander Graf
@ 2009-11-30 18:18 ` Aurelien Jarno
2009-11-30 22:21 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-11-30 18:18 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Thu, Nov 26, 2009 at 02:23:14PM +0100, Alexander Graf wrote:
> KVM on S390x requires the virtual address space of the guest's RAM to be
> within the first 256GB.
>
While I have no problem with this patch, I am curious to know the reason
of this. No need to change anything in the patch.
> The general direction I'd like to see KVM on S390 move is that this requirement
> is losened, but for now that's what we're stuck with.
>
> So let's just hack up qemu_ram_alloc until KVM behaves nicely :-).
>
> Signed-off-by: Alexander Graf <agraf@suse.de>
> ---
> exec.c | 6 ++++++
> 1 files changed, 6 insertions(+), 0 deletions(-)
>
> diff --git a/exec.c b/exec.c
> index 076d26b..59150d0 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -2411,7 +2411,13 @@ ram_addr_t qemu_ram_alloc(ram_addr_t size)
> size = TARGET_PAGE_ALIGN(size);
> new_block = qemu_malloc(sizeof(*new_block));
>
> +#if defined(TARGET_S390) && defined(CONFIG_KVM)
> + /* XXX S390 KVM requires the topmost vma of the RAM to be < 256GB */
> + new_block->host = mmap(0x1000000, size, PROT_EXEC|PROT_READ|PROT_WRITE,
> + MAP_SHARED | MAP_ANONYMOUS, -1, 0);
> +#else
> new_block->host = qemu_vmalloc(size);
> +#endif
> #ifdef MADV_MERGEABLE
> madvise(new_block->host, size, MADV_MERGEABLE);
> #endif
> --
> 1.6.0.2
>
>
>
>
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 05/11] Allocate physical memory in low virtual address space
2009-11-30 18:18 ` Aurelien Jarno
@ 2009-11-30 22:21 ` Alexander Graf
2009-12-01 9:12 ` Carsten Otte
0 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-30 22:21 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> On Thu, Nov 26, 2009 at 02:23:14PM +0100, Alexander Graf wrote:
>> KVM on S390x requires the virtual address space of the guest's RAM to be
>> within the first 256GB.
>>
>
> While I have no problem with this patch, I am curious to know the reason
> of this. No need to change anything in the patch.
For the technical reason I hope Carsten can provide some insight.
As for the implementation reason - S390 KVM (the kernel part) doesn't do memslots properly :-).
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread
* [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (4 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 05/11] Allocate physical memory in low virtual address space Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-30 18:18 ` Aurelien Jarno
2009-11-26 13:23 ` [Qemu-devel] [PATCH 07/11] Add S390x virtio machine bus Alexander Graf
` (5 subsequent siblings)
11 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
Let's enable the basics for system emulation so we can run virtual machines
with KVM!
Signed-off-by: Alexander Graf <agraf@suse.de>
---
target-s390x/cpu.h | 153 ++++++++++++++++++++++++++++++++++++-
target-s390x/exec.h | 5 +
target-s390x/helper.c | 22 +++++
target-s390x/machine.c | 30 +++++++
4 files changed, 208 insertions(+), 2 deletions(-)
create mode 100644 default-configs/s390x-softmmu.mak
create mode 100644 target-s390x/machine.c
diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
new file mode 100644
index 0000000..e69de29
diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
index f45b00c..a74745c 100644
--- a/target-s390x/cpu.h
+++ b/target-s390x/cpu.h
@@ -30,8 +30,7 @@
#include "softfloat.h"
-#define NB_MMU_MODES 2 // guess
-#define MMU_USER_IDX 0 // guess
+#define NB_MMU_MODES 2
typedef union FPReg {
struct {
@@ -77,6 +76,15 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
}
#endif
+#define MMU_MODE0_SUFFIX _kernel
+#define MMU_MODE1_SUFFIX _user
+#define MMU_USER_IDX 1
+static inline int cpu_mmu_index (CPUState *env)
+{
+ /* XXX: Currently we don't implement virtual memory */
+ return 0;
+}
+
CPUS390XState *cpu_s390x_init(const char *cpu_model);
int cpu_s390x_exec(CPUS390XState *s);
void cpu_s390x_close(CPUS390XState *s);
@@ -92,6 +100,13 @@ int cpu_s390x_handle_mmu_fault (CPUS390XState *env, target_ulong address, int rw
#define TARGET_PAGE_BITS 12
+#ifndef CONFIG_USER_ONLY
+extern int s390_virtio_hypercall(CPUState *env);
+extern void kvm_s390_virtio_irq(CPUState *env, int config_change, uint64_t token);
+extern CPUState *s390_cpu_addr2state(uint16_t cpu_addr);
+#endif
+
+
#define cpu_init cpu_s390x_init
#define cpu_exec cpu_s390x_exec
#define cpu_gen_code cpu_s390x_gen_code
@@ -116,4 +131,138 @@ static inline void cpu_get_tb_cpu_state(CPUState* env, target_ulong *pc,
*cs_base = 0;
*flags = env->psw.mask; // guess
}
+
+/* Program Status Word. */
+#define S390_PSWM_REGNUM 0
+#define S390_PSWA_REGNUM 1
+/* General Purpose Registers. */
+#define S390_R0_REGNUM 2
+#define S390_R1_REGNUM 3
+#define S390_R2_REGNUM 4
+#define S390_R3_REGNUM 5
+#define S390_R4_REGNUM 6
+#define S390_R5_REGNUM 7
+#define S390_R6_REGNUM 8
+#define S390_R7_REGNUM 9
+#define S390_R8_REGNUM 10
+#define S390_R9_REGNUM 11
+#define S390_R10_REGNUM 12
+#define S390_R11_REGNUM 13
+#define S390_R12_REGNUM 14
+#define S390_R13_REGNUM 15
+#define S390_R14_REGNUM 16
+#define S390_R15_REGNUM 17
+/* Access Registers. */
+#define S390_A0_REGNUM 18
+#define S390_A1_REGNUM 19
+#define S390_A2_REGNUM 20
+#define S390_A3_REGNUM 21
+#define S390_A4_REGNUM 22
+#define S390_A5_REGNUM 23
+#define S390_A6_REGNUM 24
+#define S390_A7_REGNUM 25
+#define S390_A8_REGNUM 26
+#define S390_A9_REGNUM 27
+#define S390_A10_REGNUM 28
+#define S390_A11_REGNUM 29
+#define S390_A12_REGNUM 30
+#define S390_A13_REGNUM 31
+#define S390_A14_REGNUM 32
+#define S390_A15_REGNUM 33
+/* Floating Point Control Word. */
+#define S390_FPC_REGNUM 34
+/* Floating Point Registers. */
+#define S390_F0_REGNUM 35
+#define S390_F1_REGNUM 36
+#define S390_F2_REGNUM 37
+#define S390_F3_REGNUM 38
+#define S390_F4_REGNUM 39
+#define S390_F5_REGNUM 40
+#define S390_F6_REGNUM 41
+#define S390_F7_REGNUM 42
+#define S390_F8_REGNUM 43
+#define S390_F9_REGNUM 44
+#define S390_F10_REGNUM 45
+#define S390_F11_REGNUM 46
+#define S390_F12_REGNUM 47
+#define S390_F13_REGNUM 48
+#define S390_F14_REGNUM 49
+#define S390_F15_REGNUM 50
+/* Total. */
+#define S390_NUM_REGS 51
+
+/* Pseudo registers -- PC and condition code. */
+#define S390_PC_REGNUM S390_NUM_REGS
+#define S390_CC_REGNUM (S390_NUM_REGS+1)
+#define S390_NUM_PSEUDO_REGS 2
+#define S390_NUM_TOTAL_REGS (S390_NUM_REGS+2)
+
+
+
+/* Program Status Word. */
+#define S390_PSWM_REGNUM 0
+#define S390_PSWA_REGNUM 1
+/* General Purpose Registers. */
+#define S390_R0_REGNUM 2
+#define S390_R1_REGNUM 3
+#define S390_R2_REGNUM 4
+#define S390_R3_REGNUM 5
+#define S390_R4_REGNUM 6
+#define S390_R5_REGNUM 7
+#define S390_R6_REGNUM 8
+#define S390_R7_REGNUM 9
+#define S390_R8_REGNUM 10
+#define S390_R9_REGNUM 11
+#define S390_R10_REGNUM 12
+#define S390_R11_REGNUM 13
+#define S390_R12_REGNUM 14
+#define S390_R13_REGNUM 15
+#define S390_R14_REGNUM 16
+#define S390_R15_REGNUM 17
+/* Access Registers. */
+#define S390_A0_REGNUM 18
+#define S390_A1_REGNUM 19
+#define S390_A2_REGNUM 20
+#define S390_A3_REGNUM 21
+#define S390_A4_REGNUM 22
+#define S390_A5_REGNUM 23
+#define S390_A6_REGNUM 24
+#define S390_A7_REGNUM 25
+#define S390_A8_REGNUM 26
+#define S390_A9_REGNUM 27
+#define S390_A10_REGNUM 28
+#define S390_A11_REGNUM 29
+#define S390_A12_REGNUM 30
+#define S390_A13_REGNUM 31
+#define S390_A14_REGNUM 32
+#define S390_A15_REGNUM 33
+/* Floating Point Control Word. */
+#define S390_FPC_REGNUM 34
+/* Floating Point Registers. */
+#define S390_F0_REGNUM 35
+#define S390_F1_REGNUM 36
+#define S390_F2_REGNUM 37
+#define S390_F3_REGNUM 38
+#define S390_F4_REGNUM 39
+#define S390_F5_REGNUM 40
+#define S390_F6_REGNUM 41
+#define S390_F7_REGNUM 42
+#define S390_F8_REGNUM 43
+#define S390_F9_REGNUM 44
+#define S390_F10_REGNUM 45
+#define S390_F11_REGNUM 46
+#define S390_F12_REGNUM 47
+#define S390_F13_REGNUM 48
+#define S390_F14_REGNUM 49
+#define S390_F15_REGNUM 50
+/* Total. */
+#define S390_NUM_REGS 51
+
+/* Pseudo registers -- PC and condition code. */
+#define S390_PC_REGNUM S390_NUM_REGS
+#define S390_CC_REGNUM (S390_NUM_REGS+1)
+#define S390_NUM_PSEUDO_REGS 2
+#define S390_NUM_TOTAL_REGS (S390_NUM_REGS+2)
+
+
#endif
diff --git a/target-s390x/exec.h b/target-s390x/exec.h
index 5198359..13dc7dd 100644
--- a/target-s390x/exec.h
+++ b/target-s390x/exec.h
@@ -22,9 +22,14 @@
register struct CPUS390XState *env asm(AREG0);
+#include "config.h"
#include "cpu.h"
#include "exec-all.h"
+#if !defined(CONFIG_USER_ONLY)
+#include "softmmu_exec.h"
+#endif /* !defined(CONFIG_USER_ONLY) */
+
static inline int cpu_has_work(CPUState *env)
{
return env->interrupt_request & CPU_INTERRUPT_HARD; // guess
diff --git a/target-s390x/helper.c b/target-s390x/helper.c
index 0e222e3..f4c4e04 100644
--- a/target-s390x/helper.c
+++ b/target-s390x/helper.c
@@ -27,6 +27,9 @@
#include "gdbstub.h"
#include "qemu-common.h"
+#include <linux/kvm.h>
+#include "kvm.h"
+
CPUS390XState *cpu_s390x_init(const char *cpu_model)
{
CPUS390XState *env;
@@ -60,3 +63,22 @@ void cpu_reset(CPUS390XState *env)
/* FIXME: reset vector? */
tlb_flush(env, 1);
}
+
+#ifndef CONFIG_USER_ONLY
+
+int cpu_s390x_handle_mmu_fault (CPUState *env, target_ulong address, int rw,
+ int mmu_idx, int is_softmmu)
+{
+ target_ulong phys;
+ int prot;
+
+ /* XXX: implement mmu */
+
+ phys = address;
+ prot = PAGE_READ | PAGE_WRITE;
+
+ return tlb_set_page(env, address & TARGET_PAGE_MASK,
+ phys & TARGET_PAGE_MASK, prot,
+ mmu_idx, is_softmmu);
+}
+#endif /* CONFIG_USER_ONLY */
diff --git a/target-s390x/machine.c b/target-s390x/machine.c
new file mode 100644
index 0000000..3e79be6
--- /dev/null
+++ b/target-s390x/machine.c
@@ -0,0 +1,30 @@
+/*
+ * QEMU S390x machine definitions
+ *
+ * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "hw/hw.h"
+#include "hw/boards.h"
+
+void cpu_save(QEMUFile *f, void *opaque)
+{
+}
+
+int cpu_load(QEMUFile *f, void *opaque, int version_id)
+{
+ return 0;
+}
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-11-26 13:23 ` [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation Alexander Graf
@ 2009-11-30 18:18 ` Aurelien Jarno
2009-11-30 22:19 ` Alexander Graf
2009-12-01 9:17 ` Carsten Otte
0 siblings, 2 replies; 49+ messages in thread
From: Aurelien Jarno @ 2009-11-30 18:18 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Thu, Nov 26, 2009 at 02:23:15PM +0100, Alexander Graf wrote:
> Let's enable the basics for system emulation so we can run virtual machines
> with KVM!
I don't really understand while this whole patch is not merged in patch
number 1. Otherwise, please find the comments below.
> Signed-off-by: Alexander Graf <agraf@suse.de>
> ---
> target-s390x/cpu.h | 153 ++++++++++++++++++++++++++++++++++++-
> target-s390x/exec.h | 5 +
> target-s390x/helper.c | 22 +++++
> target-s390x/machine.c | 30 +++++++
> 4 files changed, 208 insertions(+), 2 deletions(-)
> create mode 100644 default-configs/s390x-softmmu.mak
> create mode 100644 target-s390x/machine.c
>
> diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
> new file mode 100644
> index 0000000..e69de29
> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
> index f45b00c..a74745c 100644
> --- a/target-s390x/cpu.h
> +++ b/target-s390x/cpu.h
> @@ -30,8 +30,7 @@
>
> #include "softfloat.h"
>
> -#define NB_MMU_MODES 2 // guess
> -#define MMU_USER_IDX 0 // guess
> +#define NB_MMU_MODES 2
>
> typedef union FPReg {
> struct {
> @@ -77,6 +76,15 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
> }
> #endif
>
> +#define MMU_MODE0_SUFFIX _kernel
> +#define MMU_MODE1_SUFFIX _user
> +#define MMU_USER_IDX 1
> +static inline int cpu_mmu_index (CPUState *env)
> +{
> + /* XXX: Currently we don't implement virtual memory */
> + return 0;
Is it correct? It means that memory access will aways be kernel memory
accesses. IIRC, even with KVM enabled, softmmu accesses are possible in
some cases (devices ?).
> +}
> +
> CPUS390XState *cpu_s390x_init(const char *cpu_model);
> int cpu_s390x_exec(CPUS390XState *s);
> void cpu_s390x_close(CPUS390XState *s);
> @@ -92,6 +100,13 @@ int cpu_s390x_handle_mmu_fault (CPUS390XState *env, target_ulong address, int rw
>
> #define TARGET_PAGE_BITS 12
>
> +#ifndef CONFIG_USER_ONLY
> +extern int s390_virtio_hypercall(CPUState *env);
> +extern void kvm_s390_virtio_irq(CPUState *env, int config_change, uint64_t token);
> +extern CPUState *s390_cpu_addr2state(uint16_t cpu_addr);
> +#endif
> +
> +
> #define cpu_init cpu_s390x_init
> #define cpu_exec cpu_s390x_exec
> #define cpu_gen_code cpu_s390x_gen_code
> @@ -116,4 +131,138 @@ static inline void cpu_get_tb_cpu_state(CPUState* env, target_ulong *pc,
> *cs_base = 0;
> *flags = env->psw.mask; // guess
> }
> +
> +/* Program Status Word. */
> +#define S390_PSWM_REGNUM 0
> +#define S390_PSWA_REGNUM 1
> +/* General Purpose Registers. */
> +#define S390_R0_REGNUM 2
> +#define S390_R1_REGNUM 3
> +#define S390_R2_REGNUM 4
> +#define S390_R3_REGNUM 5
> +#define S390_R4_REGNUM 6
> +#define S390_R5_REGNUM 7
> +#define S390_R6_REGNUM 8
> +#define S390_R7_REGNUM 9
> +#define S390_R8_REGNUM 10
> +#define S390_R9_REGNUM 11
> +#define S390_R10_REGNUM 12
> +#define S390_R11_REGNUM 13
> +#define S390_R12_REGNUM 14
> +#define S390_R13_REGNUM 15
> +#define S390_R14_REGNUM 16
> +#define S390_R15_REGNUM 17
> +/* Access Registers. */
> +#define S390_A0_REGNUM 18
> +#define S390_A1_REGNUM 19
> +#define S390_A2_REGNUM 20
> +#define S390_A3_REGNUM 21
> +#define S390_A4_REGNUM 22
> +#define S390_A5_REGNUM 23
> +#define S390_A6_REGNUM 24
> +#define S390_A7_REGNUM 25
> +#define S390_A8_REGNUM 26
> +#define S390_A9_REGNUM 27
> +#define S390_A10_REGNUM 28
> +#define S390_A11_REGNUM 29
> +#define S390_A12_REGNUM 30
> +#define S390_A13_REGNUM 31
> +#define S390_A14_REGNUM 32
> +#define S390_A15_REGNUM 33
> +/* Floating Point Control Word. */
> +#define S390_FPC_REGNUM 34
> +/* Floating Point Registers. */
> +#define S390_F0_REGNUM 35
> +#define S390_F1_REGNUM 36
> +#define S390_F2_REGNUM 37
> +#define S390_F3_REGNUM 38
> +#define S390_F4_REGNUM 39
> +#define S390_F5_REGNUM 40
> +#define S390_F6_REGNUM 41
> +#define S390_F7_REGNUM 42
> +#define S390_F8_REGNUM 43
> +#define S390_F9_REGNUM 44
> +#define S390_F10_REGNUM 45
> +#define S390_F11_REGNUM 46
> +#define S390_F12_REGNUM 47
> +#define S390_F13_REGNUM 48
> +#define S390_F14_REGNUM 49
> +#define S390_F15_REGNUM 50
> +/* Total. */
> +#define S390_NUM_REGS 51
> +
> +/* Pseudo registers -- PC and condition code. */
> +#define S390_PC_REGNUM S390_NUM_REGS
> +#define S390_CC_REGNUM (S390_NUM_REGS+1)
> +#define S390_NUM_PSEUDO_REGS 2
> +#define S390_NUM_TOTAL_REGS (S390_NUM_REGS+2)
> +
> +
> +
> +/* Program Status Word. */
> +#define S390_PSWM_REGNUM 0
> +#define S390_PSWA_REGNUM 1
> +/* General Purpose Registers. */
> +#define S390_R0_REGNUM 2
> +#define S390_R1_REGNUM 3
> +#define S390_R2_REGNUM 4
> +#define S390_R3_REGNUM 5
> +#define S390_R4_REGNUM 6
> +#define S390_R5_REGNUM 7
> +#define S390_R6_REGNUM 8
> +#define S390_R7_REGNUM 9
> +#define S390_R8_REGNUM 10
> +#define S390_R9_REGNUM 11
> +#define S390_R10_REGNUM 12
> +#define S390_R11_REGNUM 13
> +#define S390_R12_REGNUM 14
> +#define S390_R13_REGNUM 15
> +#define S390_R14_REGNUM 16
> +#define S390_R15_REGNUM 17
> +/* Access Registers. */
> +#define S390_A0_REGNUM 18
> +#define S390_A1_REGNUM 19
> +#define S390_A2_REGNUM 20
> +#define S390_A3_REGNUM 21
> +#define S390_A4_REGNUM 22
> +#define S390_A5_REGNUM 23
> +#define S390_A6_REGNUM 24
> +#define S390_A7_REGNUM 25
> +#define S390_A8_REGNUM 26
> +#define S390_A9_REGNUM 27
> +#define S390_A10_REGNUM 28
> +#define S390_A11_REGNUM 29
> +#define S390_A12_REGNUM 30
> +#define S390_A13_REGNUM 31
> +#define S390_A14_REGNUM 32
> +#define S390_A15_REGNUM 33
> +/* Floating Point Control Word. */
> +#define S390_FPC_REGNUM 34
> +/* Floating Point Registers. */
> +#define S390_F0_REGNUM 35
> +#define S390_F1_REGNUM 36
> +#define S390_F2_REGNUM 37
> +#define S390_F3_REGNUM 38
> +#define S390_F4_REGNUM 39
> +#define S390_F5_REGNUM 40
> +#define S390_F6_REGNUM 41
> +#define S390_F7_REGNUM 42
> +#define S390_F8_REGNUM 43
> +#define S390_F9_REGNUM 44
> +#define S390_F10_REGNUM 45
> +#define S390_F11_REGNUM 46
> +#define S390_F12_REGNUM 47
> +#define S390_F13_REGNUM 48
> +#define S390_F14_REGNUM 49
> +#define S390_F15_REGNUM 50
> +/* Total. */
> +#define S390_NUM_REGS 51
> +
> +/* Pseudo registers -- PC and condition code. */
> +#define S390_PC_REGNUM S390_NUM_REGS
> +#define S390_CC_REGNUM (S390_NUM_REGS+1)
> +#define S390_NUM_PSEUDO_REGS 2
> +#define S390_NUM_TOTAL_REGS (S390_NUM_REGS+2)
> +
> +
> #endif
> diff --git a/target-s390x/exec.h b/target-s390x/exec.h
> index 5198359..13dc7dd 100644
> --- a/target-s390x/exec.h
> +++ b/target-s390x/exec.h
> @@ -22,9 +22,14 @@
>
> register struct CPUS390XState *env asm(AREG0);
>
> +#include "config.h"
> #include "cpu.h"
> #include "exec-all.h"
>
> +#if !defined(CONFIG_USER_ONLY)
> +#include "softmmu_exec.h"
> +#endif /* !defined(CONFIG_USER_ONLY) */
> +
> static inline int cpu_has_work(CPUState *env)
> {
> return env->interrupt_request & CPU_INTERRUPT_HARD; // guess
> diff --git a/target-s390x/helper.c b/target-s390x/helper.c
> index 0e222e3..f4c4e04 100644
> --- a/target-s390x/helper.c
> +++ b/target-s390x/helper.c
> @@ -27,6 +27,9 @@
> #include "gdbstub.h"
> #include "qemu-common.h"
>
> +#include <linux/kvm.h>
> +#include "kvm.h"
> +
> CPUS390XState *cpu_s390x_init(const char *cpu_model)
> {
> CPUS390XState *env;
> @@ -60,3 +63,22 @@ void cpu_reset(CPUS390XState *env)
> /* FIXME: reset vector? */
> tlb_flush(env, 1);
> }
> +
> +#ifndef CONFIG_USER_ONLY
> +
> +int cpu_s390x_handle_mmu_fault (CPUState *env, target_ulong address, int rw,
> + int mmu_idx, int is_softmmu)
> +{
> + target_ulong phys;
> + int prot;
> +
> + /* XXX: implement mmu */
> +
> + phys = address;
> + prot = PAGE_READ | PAGE_WRITE;
> +
> + return tlb_set_page(env, address & TARGET_PAGE_MASK,
> + phys & TARGET_PAGE_MASK, prot,
> + mmu_idx, is_softmmu);
Same as previous comment, if softmmu accesses are possible, this
function should be written correctly.
> +}
> +#endif /* CONFIG_USER_ONLY */
> diff --git a/target-s390x/machine.c b/target-s390x/machine.c
> new file mode 100644
> index 0000000..3e79be6
> --- /dev/null
> +++ b/target-s390x/machine.c
> @@ -0,0 +1,30 @@
> +/*
> + * QEMU S390x machine definitions
> + *
> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
/> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include "hw/hw.h"
> +#include "hw/boards.h"
> +
> +void cpu_save(QEMUFile *f, void *opaque)
> +{
> +}
> +
> +int cpu_load(QEMUFile *f, void *opaque, int version_id)
> +{
> + return 0;
> +}
> --
> 1.6.0.2
>
>
>
>
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-11-30 18:18 ` Aurelien Jarno
@ 2009-11-30 22:19 ` Alexander Graf
2009-12-02 8:09 ` Aurelien Jarno
2009-12-01 9:17 ` Carsten Otte
1 sibling, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-30 22:19 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> On Thu, Nov 26, 2009 at 02:23:15PM +0100, Alexander Graf wrote:
>> Let's enable the basics for system emulation so we can run virtual machines
>> with KVM!
>
> I don't really understand while this whole patch is not merged in patch
> number 1. Otherwise, please find the comments below.
Historical reasons. To keep Uli's stripped down version separate from my code.
>
>> Signed-off-by: Alexander Graf <agraf@suse.de>
>> ---
>> target-s390x/cpu.h | 153 ++++++++++++++++++++++++++++++++++++-
>> target-s390x/exec.h | 5 +
>> target-s390x/helper.c | 22 +++++
>> target-s390x/machine.c | 30 +++++++
>> 4 files changed, 208 insertions(+), 2 deletions(-)
>> create mode 100644 default-configs/s390x-softmmu.mak
>> create mode 100644 target-s390x/machine.c
>>
>> diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
>> new file mode 100644
>> index 0000000..e69de29
>> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
>> index f45b00c..a74745c 100644
>> --- a/target-s390x/cpu.h
>> +++ b/target-s390x/cpu.h
>> @@ -30,8 +30,7 @@
>>
>> #include "softfloat.h"
>>
>> -#define NB_MMU_MODES 2 // guess
>> -#define MMU_USER_IDX 0 // guess
>> +#define NB_MMU_MODES 2
>>
>> typedef union FPReg {
>> struct {
>> @@ -77,6 +76,15 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
>> }
>> #endif
>>
>> +#define MMU_MODE0_SUFFIX _kernel
>> +#define MMU_MODE1_SUFFIX _user
>> +#define MMU_USER_IDX 1
>> +static inline int cpu_mmu_index (CPUState *env)
>> +{
>> + /* XXX: Currently we don't implement virtual memory */
>> + return 0;
>
> Is it correct? It means that memory access will aways be kernel memory
> accesses. IIRC, even with KVM enabled, softmmu accesses are possible in
> some cases (devices ?).
I can't imagine any hardware using the CPU's MMU to write to RAM. That's what IOMMUs are for.
The only 2 consumers are:
1) tcg
2) gdb / monitor
With 2) being broken, because we can't resolve virtual addresses to physical addresses. But that won't change until someone implements the softmmu emulation target for real.
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-11-30 22:19 ` Alexander Graf
@ 2009-12-02 8:09 ` Aurelien Jarno
2009-12-02 8:27 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-12-02 8:09 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Mon, Nov 30, 2009 at 11:19:06PM +0100, Alexander Graf wrote:
>
> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>
> > On Thu, Nov 26, 2009 at 02:23:15PM +0100, Alexander Graf wrote:
> >> Let's enable the basics for system emulation so we can run virtual machines
> >> with KVM!
> >
> > I don't really understand while this whole patch is not merged in patch
> > number 1. Otherwise, please find the comments below.
>
> Historical reasons. To keep Uli's stripped down version separate from my code.
>
> >
> >> Signed-off-by: Alexander Graf <agraf@suse.de>
> >> ---
> >> target-s390x/cpu.h | 153 ++++++++++++++++++++++++++++++++++++-
> >> target-s390x/exec.h | 5 +
> >> target-s390x/helper.c | 22 +++++
> >> target-s390x/machine.c | 30 +++++++
> >> 4 files changed, 208 insertions(+), 2 deletions(-)
> >> create mode 100644 default-configs/s390x-softmmu.mak
> >> create mode 100644 target-s390x/machine.c
> >>
> >> diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
> >> new file mode 100644
> >> index 0000000..e69de29
> >> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
> >> index f45b00c..a74745c 100644
> >> --- a/target-s390x/cpu.h
> >> +++ b/target-s390x/cpu.h
> >> @@ -30,8 +30,7 @@
> >>
> >> #include "softfloat.h"
> >>
> >> -#define NB_MMU_MODES 2 // guess
> >> -#define MMU_USER_IDX 0 // guess
> >> +#define NB_MMU_MODES 2
> >>
> >> typedef union FPReg {
> >> struct {
> >> @@ -77,6 +76,15 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
> >> }
> >> #endif
> >>
> >> +#define MMU_MODE0_SUFFIX _kernel
> >> +#define MMU_MODE1_SUFFIX _user
> >> +#define MMU_USER_IDX 1
> >> +static inline int cpu_mmu_index (CPUState *env)
> >> +{
> >> + /* XXX: Currently we don't implement virtual memory */
> >> + return 0;
> >
> > Is it correct? It means that memory access will aways be kernel memory
> > accesses. IIRC, even with KVM enabled, softmmu accesses are possible in
> > some cases (devices ?).
>
> I can't imagine any hardware using the CPU's MMU to write to RAM. That's what IOMMUs are for.
>
> The only 2 consumers are:
>
> 1) tcg
> 2) gdb / monitor
>
> With 2) being broken, because we can't resolve virtual addresses to physical addresses. But that won't change until someone implements the softmmu emulation target for real.
If it is sure it is never used, I would prefer to see an abort().
Otherwise it's fine.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-12-02 8:09 ` Aurelien Jarno
@ 2009-12-02 8:27 ` Alexander Graf
2009-12-02 8:37 ` Aurelien Jarno
0 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-12-02 8:27 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 02.12.2009, at 09:09, Aurelien Jarno wrote:
> On Mon, Nov 30, 2009 at 11:19:06PM +0100, Alexander Graf wrote:
>>
>> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>>
>>> On Thu, Nov 26, 2009 at 02:23:15PM +0100, Alexander Graf wrote:
>>>> Let's enable the basics for system emulation so we can run virtual machines
>>>> with KVM!
>>>
>>> I don't really understand while this whole patch is not merged in patch
>>> number 1. Otherwise, please find the comments below.
>>
>> Historical reasons. To keep Uli's stripped down version separate from my code.
>>
>>>
>>>> Signed-off-by: Alexander Graf <agraf@suse.de>
>>>> ---
>>>> target-s390x/cpu.h | 153 ++++++++++++++++++++++++++++++++++++-
>>>> target-s390x/exec.h | 5 +
>>>> target-s390x/helper.c | 22 +++++
>>>> target-s390x/machine.c | 30 +++++++
>>>> 4 files changed, 208 insertions(+), 2 deletions(-)
>>>> create mode 100644 default-configs/s390x-softmmu.mak
>>>> create mode 100644 target-s390x/machine.c
>>>>
>>>> diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
>>>> new file mode 100644
>>>> index 0000000..e69de29
>>>> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
>>>> index f45b00c..a74745c 100644
>>>> --- a/target-s390x/cpu.h
>>>> +++ b/target-s390x/cpu.h
>>>> @@ -30,8 +30,7 @@
>>>>
>>>> #include "softfloat.h"
>>>>
>>>> -#define NB_MMU_MODES 2 // guess
>>>> -#define MMU_USER_IDX 0 // guess
>>>> +#define NB_MMU_MODES 2
>>>>
>>>> typedef union FPReg {
>>>> struct {
>>>> @@ -77,6 +76,15 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
>>>> }
>>>> #endif
>>>>
>>>> +#define MMU_MODE0_SUFFIX _kernel
>>>> +#define MMU_MODE1_SUFFIX _user
>>>> +#define MMU_USER_IDX 1
>>>> +static inline int cpu_mmu_index (CPUState *env)
>>>> +{
>>>> + /* XXX: Currently we don't implement virtual memory */
>>>> + return 0;
>>>
>>> Is it correct? It means that memory access will aways be kernel memory
>>> accesses. IIRC, even with KVM enabled, softmmu accesses are possible in
>>> some cases (devices ?).
>>
>> I can't imagine any hardware using the CPU's MMU to write to RAM. That's what IOMMUs are for.
>>
>> The only 2 consumers are:
>>
>> 1) tcg
>> 2) gdb / monitor
>>
>> With 2) being broken, because we can't resolve virtual addresses to physical addresses. But that won't change until someone implements the softmmu emulation target for real.
>
> If it is sure it is never used, I would prefer to see an abort().
> Otherwise it's fine.
I don't think I understand where you want to put the abort().
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-12-02 8:27 ` Alexander Graf
@ 2009-12-02 8:37 ` Aurelien Jarno
2009-12-02 8:41 ` Alexander Graf
0 siblings, 1 reply; 49+ messages in thread
From: Aurelien Jarno @ 2009-12-02 8:37 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Wed, Dec 02, 2009 at 09:27:21AM +0100, Alexander Graf wrote:
>
> On 02.12.2009, at 09:09, Aurelien Jarno wrote:
>
> > On Mon, Nov 30, 2009 at 11:19:06PM +0100, Alexander Graf wrote:
> >>
> >> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
> >>
> >>> On Thu, Nov 26, 2009 at 02:23:15PM +0100, Alexander Graf wrote:
> >>>> Let's enable the basics for system emulation so we can run virtual machines
> >>>> with KVM!
> >>>
> >>> I don't really understand while this whole patch is not merged in patch
> >>> number 1. Otherwise, please find the comments below.
> >>
> >> Historical reasons. To keep Uli's stripped down version separate from my code.
> >>
> >>>
> >>>> Signed-off-by: Alexander Graf <agraf@suse.de>
> >>>> ---
> >>>> target-s390x/cpu.h | 153 ++++++++++++++++++++++++++++++++++++-
> >>>> target-s390x/exec.h | 5 +
> >>>> target-s390x/helper.c | 22 +++++
> >>>> target-s390x/machine.c | 30 +++++++
> >>>> 4 files changed, 208 insertions(+), 2 deletions(-)
> >>>> create mode 100644 default-configs/s390x-softmmu.mak
> >>>> create mode 100644 target-s390x/machine.c
> >>>>
> >>>> diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
> >>>> new file mode 100644
> >>>> index 0000000..e69de29
> >>>> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
> >>>> index f45b00c..a74745c 100644
> >>>> --- a/target-s390x/cpu.h
> >>>> +++ b/target-s390x/cpu.h
> >>>> @@ -30,8 +30,7 @@
> >>>>
> >>>> #include "softfloat.h"
> >>>>
> >>>> -#define NB_MMU_MODES 2 // guess
> >>>> -#define MMU_USER_IDX 0 // guess
> >>>> +#define NB_MMU_MODES 2
> >>>>
> >>>> typedef union FPReg {
> >>>> struct {
> >>>> @@ -77,6 +76,15 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
> >>>> }
> >>>> #endif
> >>>>
> >>>> +#define MMU_MODE0_SUFFIX _kernel
> >>>> +#define MMU_MODE1_SUFFIX _user
> >>>> +#define MMU_USER_IDX 1
> >>>> +static inline int cpu_mmu_index (CPUState *env)
> >>>> +{
> >>>> + /* XXX: Currently we don't implement virtual memory */
> >>>> + return 0;
> >>>
> >>> Is it correct? It means that memory access will aways be kernel memory
> >>> accesses. IIRC, even with KVM enabled, softmmu accesses are possible in
> >>> some cases (devices ?).
> >>
> >> I can't imagine any hardware using the CPU's MMU to write to RAM. That's what IOMMUs are for.
> >>
> >> The only 2 consumers are:
> >>
> >> 1) tcg
> >> 2) gdb / monitor
> >>
> >> With 2) being broken, because we can't resolve virtual addresses to physical addresses. But that won't change until someone implements the softmmu emulation target for real.
> >
> > If it is sure it is never used, I would prefer to see an abort().
> > Otherwise it's fine.
>
> I don't think I understand where you want to put the abort().
>
In inline cpu_mmu_index(), just before the return, to make sure this
function is never called, as it is clearly wrong.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-12-02 8:37 ` Aurelien Jarno
@ 2009-12-02 8:41 ` Alexander Graf
0 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-12-02 8:41 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, qemu-devel
On 02.12.2009, at 09:37, Aurelien Jarno wrote:
> On Wed, Dec 02, 2009 at 09:27:21AM +0100, Alexander Graf wrote:
>>
>> On 02.12.2009, at 09:09, Aurelien Jarno wrote:
>>
>>> On Mon, Nov 30, 2009 at 11:19:06PM +0100, Alexander Graf wrote:
>>>>
>>>> On 30.11.2009, at 19:18, Aurelien Jarno wrote:
>>>>
>>>>> On Thu, Nov 26, 2009 at 02:23:15PM +0100, Alexander Graf wrote:
>>>>>> Let's enable the basics for system emulation so we can run virtual machines
>>>>>> with KVM!
>>>>>
>>>>> I don't really understand while this whole patch is not merged in patch
>>>>> number 1. Otherwise, please find the comments below.
>>>>
>>>> Historical reasons. To keep Uli's stripped down version separate from my code.
>>>>
>>>>>
>>>>>> Signed-off-by: Alexander Graf <agraf@suse.de>
>>>>>> ---
>>>>>> target-s390x/cpu.h | 153 ++++++++++++++++++++++++++++++++++++-
>>>>>> target-s390x/exec.h | 5 +
>>>>>> target-s390x/helper.c | 22 +++++
>>>>>> target-s390x/machine.c | 30 +++++++
>>>>>> 4 files changed, 208 insertions(+), 2 deletions(-)
>>>>>> create mode 100644 default-configs/s390x-softmmu.mak
>>>>>> create mode 100644 target-s390x/machine.c
>>>>>>
>>>>>> diff --git a/default-configs/s390x-softmmu.mak b/default-configs/s390x-softmmu.mak
>>>>>> new file mode 100644
>>>>>> index 0000000..e69de29
>>>>>> diff --git a/target-s390x/cpu.h b/target-s390x/cpu.h
>>>>>> index f45b00c..a74745c 100644
>>>>>> --- a/target-s390x/cpu.h
>>>>>> +++ b/target-s390x/cpu.h
>>>>>> @@ -30,8 +30,7 @@
>>>>>>
>>>>>> #include "softfloat.h"
>>>>>>
>>>>>> -#define NB_MMU_MODES 2 // guess
>>>>>> -#define MMU_USER_IDX 0 // guess
>>>>>> +#define NB_MMU_MODES 2
>>>>>>
>>>>>> typedef union FPReg {
>>>>>> struct {
>>>>>> @@ -77,6 +76,15 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
>>>>>> }
>>>>>> #endif
>>>>>>
>>>>>> +#define MMU_MODE0_SUFFIX _kernel
>>>>>> +#define MMU_MODE1_SUFFIX _user
>>>>>> +#define MMU_USER_IDX 1
>>>>>> +static inline int cpu_mmu_index (CPUState *env)
>>>>>> +{
>>>>>> + /* XXX: Currently we don't implement virtual memory */
>>>>>> + return 0;
>>>>>
>>>>> Is it correct? It means that memory access will aways be kernel memory
>>>>> accesses. IIRC, even with KVM enabled, softmmu accesses are possible in
>>>>> some cases (devices ?).
>>>>
>>>> I can't imagine any hardware using the CPU's MMU to write to RAM. That's what IOMMUs are for.
>>>>
>>>> The only 2 consumers are:
>>>>
>>>> 1) tcg
>>>> 2) gdb / monitor
>>>>
>>>> With 2) being broken, because we can't resolve virtual addresses to physical addresses. But that won't change until someone implements the softmmu emulation target for real.
>>>
>>> If it is sure it is never used, I would prefer to see an abort().
>>> Otherwise it's fine.
>>
>> I don't think I understand where you want to put the abort().
>>
>
> In inline cpu_mmu_index(), just before the return, to make sure this
> function is never called, as it is clearly wrong.
It's just always saying we're in kernel mode. I don't see where that's wrong. There's no logic to implement modes, so that's the only reasonable thing to do.
Also, it does get called. That's what I mean with the target 2). When you're in the monitor and do x /i $pc, you end up calling that function. While it's not great to only have linear mapped memory here, it's a lot better than having no reply or, even worse, killing the VM.
Alex
^ permalink raw reply [flat|nested] 49+ messages in thread
* Re: [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation
2009-11-30 18:18 ` Aurelien Jarno
2009-11-30 22:19 ` Alexander Graf
@ 2009-12-01 9:17 ` Carsten Otte
1 sibling, 0 replies; 49+ messages in thread
From: Carsten Otte @ 2009-12-01 9:17 UTC (permalink / raw)
To: Aurelien Jarno; +Cc: Carsten Otte, Alexander Graf, qemu-devel
Aurelien Jarno wrote:
>> +static inline int cpu_mmu_index (CPUState *env)
>> +{
>> + /* XXX: Currently we don't implement virtual memory */
>> + return 0;
>
> Is it correct? It means that memory access will aways be kernel memory
> accesses. IIRC, even with KVM enabled, softmmu accesses are possible in
> some cases (devices ?).
On s390, we don't have memory mapped I/O. KVM is virtio only, and even
if we would add channel devices we would not require softmmu emulation
here. S390 does'nt have pci, usb or alike.
^ permalink raw reply [flat|nested] 49+ messages in thread
* [Qemu-devel] [PATCH 07/11] Add S390x virtio machine bus
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (5 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 06/11] Add support for S390x system emulation Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-26 13:23 ` [Qemu-devel] [PATCH 08/11] Add S390x virtio machine description Alexander Graf
` (4 subsequent siblings)
11 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
On S390x we don't want to go through the hassle of emulating real existing
hardware, because we don't need to for running Linux.
So let's instead implement a machine that is 100% based on VirtIO which we
fortunately implement already.
This patch implements the bus that is the groundwork for such an S390x
virtio machine.
Signed-off-by: Alexander Graf <agraf@suse.de>
---
Makefile.target | 2 +
hw/s390-virtio-bus.c | 385 ++++++++++++++++++++++++++++++++++++++++++++++++++
hw/s390-virtio-bus.h | 65 +++++++++
3 files changed, 452 insertions(+), 0 deletions(-)
create mode 100644 hw/s390-virtio-bus.c
create mode 100644 hw/s390-virtio-bus.h
diff --git a/Makefile.target b/Makefile.target
index 2985658..1619c23 100644
--- a/Makefile.target
+++ b/Makefile.target
@@ -296,6 +296,8 @@ obj-sh4-y += ide/core.o ide/mmio.o
obj-m68k-y = an5206.o mcf5206.o mcf_uart.o mcf_intc.o mcf5208.o mcf_fec.o
obj-m68k-y += m68k-semi.o dummy_m68k.o
+obj-s390x-y = s390-virtio-bus.o
+
main.o vl.o: QEMU_CFLAGS+=$(GPROF_CFLAGS)
vl.o: QEMU_CFLAGS+=$(SDL_CFLAGS)
diff --git a/hw/s390-virtio-bus.c b/hw/s390-virtio-bus.c
new file mode 100644
index 0000000..c4b6e69
--- /dev/null
+++ b/hw/s390-virtio-bus.c
@@ -0,0 +1,385 @@
+/*
+ * QEMU S390 virtio target
+ *
+ * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "hw.h"
+#include "block.h"
+#include "sysemu.h"
+#include "net.h"
+#include "boards.h"
+#include "monitor.h"
+#include "loader.h"
+#include "elf.h"
+#include "hw/virtio.h"
+#include "hw/virtio-console.h"
+#include "hw/sysbus.h"
+#include "kvm.h"
+
+#include "hw/s390-virtio-bus.h"
+
+//#define DEBUG_S390
+
+#ifdef DEBUG_S390
+#define dprintf(fmt, ...) \
+ do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
+#else
+#define dprintf(fmt, ...) \
+ do { } while (0)
+#endif
+
+struct BusInfo s390_virtio_bus_info = {
+ .name = "s390-virtio",
+ .size = sizeof(VirtIOS390Bus),
+};
+
+typedef struct {
+ DeviceInfo qdev;
+ int (*init)(VirtIOS390Device *dev);
+} VirtIOS390DeviceInfo;
+
+
+static const VirtIOBindings virtio_s390_bindings;
+
+static ram_addr_t s390_virtio_device_num_vq(VirtIOS390Device *dev);
+static void s390_virtio_device_sync(VirtIOS390Device *dev);
+
+VirtIOS390Bus *s390_virtio_bus_init(ram_addr_t *ram_size)
+{
+ VirtIOS390Bus *bus;
+ BusState *_bus;
+ DeviceState *dev;
+
+ /* Create bridge device */
+ dev = qdev_create(NULL, "s390-virtio-bridge");
+ qdev_init_nofail(dev);
+
+ /* Create bus on bridge device */
+
+ _bus = qbus_create(&s390_virtio_bus_info, dev, "s390-virtio");
+ bus = DO_UPCAST(VirtIOS390Bus, bus, _bus);
+
+ bus->dev_page = *ram_size;
+ bus->dev_offs = bus->dev_page;
+ bus->next_ring = bus->dev_page + TARGET_PAGE_SIZE;
+
+ /* Allocate RAM for VirtIO device pages (descriptors, queues, rings) */
+ *ram_size += S390_DEVICE_PAGES * TARGET_PAGE_SIZE;
+
+ return bus;
+}
+
+static int s390_virtio_device_init(VirtIOS390Device *dev, VirtIODevice *vdev)
+{
+ VirtIOS390Bus *bus;
+ int dev_len;
+
+ bus = DO_UPCAST(VirtIOS390Bus, bus, dev->qdev.parent_bus);
+ dev->vdev = vdev;
+ dev->dev_offs = bus->dev_offs;
+ dev->feat_len = sizeof(uint32_t); /* always keep 32 bits features */
+
+ dev_len = VIRTIO_DEV_OFFS_CONFIG;
+ dev_len += s390_virtio_device_num_vq(dev) * VIRTIO_VQCONFIG_LEN;
+ dev_len += dev->feat_len * 2;
+ dev_len += vdev->config_len;
+
+ bus->dev_offs += dev_len;
+
+ virtio_bind_device(vdev, &virtio_s390_bindings, dev);
+ s390_virtio_device_sync(dev);
+
+ return 0;
+}
+
+static int s390_virtio_net_init(VirtIOS390Device *dev)
+{
+ VirtIODevice *vdev;
+
+ vdev = virtio_net_init((DeviceState *)dev, &dev->nic);
+ if (!vdev)
+ return -1;
+
+ return s390_virtio_device_init(dev, vdev);
+}
+
+static int s390_virtio_blk_init(VirtIOS390Device *dev)
+{
+ VirtIODevice *vdev;
+
+ vdev = virtio_blk_init((DeviceState *)dev, dev->dinfo);
+ if (!vdev)
+ return -1;
+
+ return s390_virtio_device_init(dev, vdev);
+}
+
+static int s390_virtio_console_init(VirtIOS390Device *dev)
+{
+ VirtIOS390Bus *bus;
+ VirtIODevice *vdev;
+ int r;
+
+ bus = DO_UPCAST(VirtIOS390Bus, bus, dev->qdev.parent_bus);
+
+ vdev = virtio_console_init((DeviceState *)dev);
+ if (!vdev)
+ return -1;
+
+ r = s390_virtio_device_init(dev, vdev);
+ if (!r)
+ bus->console = dev;
+
+ return r;
+}
+
+static uint64_t s390_virtio_device_vq_token(VirtIOS390Device *dev, int vq)
+{
+ ram_addr_t token_off;
+
+ token_off = (dev->dev_offs + VIRTIO_DEV_OFFS_CONFIG) +
+ (vq * VIRTIO_VQCONFIG_LEN) +
+ VIRTIO_VQCONFIG_OFFS_TOKEN;
+
+ return ldq_phys(token_off);
+}
+
+static ram_addr_t s390_virtio_device_num_vq(VirtIOS390Device *dev)
+{
+ VirtIODevice *vdev = dev->vdev;
+ int num_vq;
+
+ for (num_vq = 0; num_vq < VIRTIO_PCI_QUEUE_MAX; num_vq++) {
+ if (!virtio_queue_get_num(vdev, num_vq))
+ break;
+ }
+
+ return num_vq;
+}
+
+static ram_addr_t s390_virtio_next_ring(VirtIOS390Bus *bus)
+{
+ ram_addr_t r = bus->next_ring;
+
+ bus->next_ring += VIRTIO_RING_LEN;
+ return r;
+}
+
+static void s390_virtio_device_sync(VirtIOS390Device *dev)
+{
+ VirtIOS390Bus *bus = DO_UPCAST(VirtIOS390Bus, bus, dev->qdev.parent_bus);
+ ram_addr_t cur_offs;
+ uint8_t num_vq;
+ int i;
+
+ virtio_reset(dev->vdev);
+
+ /* Sync dev space */
+ stb_phys(dev->dev_offs + VIRTIO_DEV_OFFS_TYPE, dev->vdev->device_id);
+
+ stb_phys(dev->dev_offs + VIRTIO_DEV_OFFS_NUM_VQ, s390_virtio_device_num_vq(dev));
+ stb_phys(dev->dev_offs + VIRTIO_DEV_OFFS_FEATURE_LEN, dev->feat_len);
+
+ stb_phys(dev->dev_offs + VIRTIO_DEV_OFFS_CONFIG_LEN, dev->vdev->config_len);
+
+ num_vq = s390_virtio_device_num_vq(dev);
+ stb_phys(dev->dev_offs + VIRTIO_DEV_OFFS_NUM_VQ, num_vq);
+
+ /* Sync virtqueues */
+ for (i = 0; i < num_vq; i++) {
+ ram_addr_t vq = (dev->dev_offs + VIRTIO_DEV_OFFS_CONFIG) +
+ (i * VIRTIO_VQCONFIG_LEN);
+ ram_addr_t vring;
+
+ vring = s390_virtio_next_ring(bus);
+ virtio_queue_set_addr(dev->vdev, i, vring);
+ virtio_queue_set_vector(dev->vdev, i, i);
+ stq_phys(vq + VIRTIO_VQCONFIG_OFFS_ADDRESS, vring);
+ stw_phys(vq + VIRTIO_VQCONFIG_OFFS_NUM, virtio_queue_get_num(dev->vdev, i));
+ }
+
+ cur_offs = dev->dev_offs;
+ cur_offs += VIRTIO_DEV_OFFS_CONFIG;
+ cur_offs += num_vq * VIRTIO_VQCONFIG_LEN;
+
+ /* Sync feature bitmap */
+ if (dev->vdev->get_features)
+ stl_phys(cur_offs, dev->vdev->get_features(dev->vdev));
+
+ dev->feat_offs = cur_offs + dev->feat_len;
+ cur_offs += dev->feat_len * 2;
+
+ /* Sync config space */
+ if (dev->vdev->get_config)
+ dev->vdev->get_config(dev->vdev, dev->vdev->config);
+
+ cpu_physical_memory_rw(cur_offs, dev->vdev->config, dev->vdev->config_len, 1);
+ cur_offs += dev->vdev->config_len;
+}
+
+void s390_virtio_device_update_status(VirtIOS390Device *dev)
+{
+ VirtIODevice *vdev = dev->vdev;
+ uint32_t features;
+
+ vdev->status = ldub_phys(dev->dev_offs + VIRTIO_DEV_OFFS_STATUS);
+
+ /* Update guest supported feature bitmap */
+
+ features = ldl_phys(dev->feat_offs);
+ if (vdev->set_features)
+ vdev->set_features(vdev, features);
+ vdev->features = features;
+}
+
+VirtIOS390Device *s390_virtio_bus_console(VirtIOS390Bus *bus)
+{
+ return bus->console;
+}
+
+/* Find a device by vring address */
+VirtIOS390Device *s390_virtio_bus_find_vring(VirtIOS390Bus *bus,
+ ram_addr_t mem,
+ int *vq_num)
+{
+ VirtIOS390Device *_dev;
+ DeviceState *dev;
+ int i;
+
+ QLIST_FOREACH(dev, &bus->bus.children, sibling) {
+ _dev = (VirtIOS390Device *)dev;
+ for(i = 0; i < VIRTIO_PCI_QUEUE_MAX; i++) {
+ if (!virtio_queue_get_addr(_dev->vdev, i))
+ break;
+ if (virtio_queue_get_addr(_dev->vdev, i) == mem) {
+ if (vq_num) *vq_num = i;
+ return _dev;
+ }
+ }
+ }
+
+ return NULL;
+}
+
+/* Find a device by device descriptor location */
+VirtIOS390Device *s390_virtio_bus_find_mem(VirtIOS390Bus *bus, ram_addr_t mem)
+{
+ VirtIOS390Device *_dev;
+ DeviceState *dev;
+
+ QLIST_FOREACH(dev, &bus->bus.children, sibling) {
+ _dev = (VirtIOS390Device *)dev;
+ if (_dev->dev_offs == mem) {
+ return _dev;
+ }
+ }
+
+ return NULL;
+}
+
+static void virtio_s390_notify(void *opaque, uint16_t vector)
+{
+ VirtIOS390Device *dev = (VirtIOS390Device*)opaque;
+ uint64_t token = s390_virtio_device_vq_token(dev, vector);
+
+ /* XXX kvm dependency! */
+ kvm_s390_virtio_irq(s390_cpu_addr2state(0), 1, token);
+}
+
+/**************** S390 Virtio Bus Device Descriptions *******************/
+
+static const VirtIOBindings virtio_s390_bindings = {
+ .notify = virtio_s390_notify,
+};
+
+static VirtIOS390DeviceInfo s390_virtio_net = {
+ .init = s390_virtio_net_init,
+ .qdev.name = "virtio-net-s390",
+ .qdev.size = sizeof(VirtIOS390Device),
+ .qdev.props = (Property[]) {
+ DEFINE_NIC_PROPERTIES(VirtIOS390Device, nic),
+ DEFINE_PROP_END_OF_LIST(),
+ },
+};
+
+static VirtIOS390DeviceInfo s390_virtio_blk = {
+ .init = s390_virtio_blk_init,
+ .qdev.name = "virtio-blk-s390",
+ .qdev.size = sizeof(VirtIOS390Device),
+ .qdev.props = (Property[]) {
+ DEFINE_PROP_DRIVE("drive", VirtIOS390Device, dinfo),
+ DEFINE_PROP_END_OF_LIST(),
+ },
+};
+
+static VirtIOS390DeviceInfo s390_virtio_console = {
+ .init = s390_virtio_console_init,
+ .qdev.name = "virtio-console-s390",
+ .qdev.size = sizeof(VirtIOS390Device),
+ .qdev.props = (Property[]) {
+ DEFINE_PROP_END_OF_LIST(),
+ },
+};
+
+static int s390_virtio_busdev_init(DeviceState *dev, DeviceInfo *info)
+{
+ VirtIOS390DeviceInfo *_info = (VirtIOS390DeviceInfo *)info;
+ VirtIOS390Device *_dev = (VirtIOS390Device *)dev;
+
+ return _info->init(_dev);
+}
+
+static void s390_virtio_bus_register_withprop(VirtIOS390DeviceInfo *info)
+{
+ info->qdev.init = s390_virtio_busdev_init;
+ info->qdev.bus_info = &s390_virtio_bus_info;
+
+ assert(info->qdev.size >= sizeof(VirtIOS390Device));
+ qdev_register(&info->qdev);
+}
+
+static void s390_virtio_register(void)
+{
+ s390_virtio_bus_register_withprop(&s390_virtio_console);
+ s390_virtio_bus_register_withprop(&s390_virtio_blk);
+ s390_virtio_bus_register_withprop(&s390_virtio_net);
+}
+device_init(s390_virtio_register);
+
+
+/***************** S390 Virtio Bus Bridge Device *******************/
+/* Only required to have the virtio bus as child in the system bus */
+
+static int s390_virtio_bridge_init(SysBusDevice *dev)
+{
+ /* nothing */
+ return 0;
+}
+
+static SysBusDeviceInfo s390_virtio_bridge_info = {
+ .init = s390_virtio_bridge_init,
+ .qdev.name = "s390-virtio-bridge",
+ .qdev.size = sizeof(SysBusDevice),
+ .qdev.no_user = 1,
+};
+
+static void s390_virtio_register_devices(void)
+{
+ sysbus_register_withprop(&s390_virtio_bridge_info);
+}
+
+device_init(s390_virtio_register_devices)
diff --git a/hw/s390-virtio-bus.h b/hw/s390-virtio-bus.h
new file mode 100644
index 0000000..e84fd64
--- /dev/null
+++ b/hw/s390-virtio-bus.h
@@ -0,0 +1,65 @@
+/*
+ * QEMU S390x VirtIO BUS definitions
+ *
+ * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#define VIRTIO_DEV_OFFS_TYPE 0 /* 8 bits */
+#define VIRTIO_DEV_OFFS_NUM_VQ 1 /* 8 bits */
+#define VIRTIO_DEV_OFFS_FEATURE_LEN 2 /* 8 bits */
+#define VIRTIO_DEV_OFFS_CONFIG_LEN 3 /* 8 bits */
+#define VIRTIO_DEV_OFFS_STATUS 4 /* 8 bits */
+#define VIRTIO_DEV_OFFS_CONFIG 5 /* dynamic */
+
+#define VIRTIO_VQCONFIG_OFFS_TOKEN 0 /* 64 bits */
+#define VIRTIO_VQCONFIG_OFFS_ADDRESS 8 /* 64 bits */
+#define VIRTIO_VQCONFIG_OFFS_NUM 16 /* 16 bits */
+#define VIRTIO_VQCONFIG_LEN 24
+
+#define VIRTIO_RING_LEN (TARGET_PAGE_SIZE * 3)
+#define S390_DEVICE_PAGES 256
+
+typedef struct VirtIOS390Device {
+ DeviceState qdev;
+ ram_addr_t dev_offs;
+ ram_addr_t feat_offs;
+ uint8_t feat_len;
+ VirtIODevice *vdev;
+ DriveInfo *dinfo;
+ NICConf nic;
+} VirtIOS390Device;
+
+typedef struct VirtIOS390Bus {
+ BusState bus;
+
+ VirtIOS390Device *console;
+ ram_addr_t dev_page;
+ ram_addr_t dev_offs;
+ ram_addr_t next_ring;
+} VirtIOS390Bus;
+
+
+extern void s390_virtio_device_update_status(VirtIOS390Device *dev);
+
+extern VirtIOS390Device *s390_virtio_bus_console(VirtIOS390Bus *bus);
+extern VirtIOS390Bus *s390_virtio_bus_init(ram_addr_t *ram_size);
+
+extern VirtIOS390Device *s390_virtio_bus_find_vring(VirtIOS390Bus *bus,
+ ram_addr_t mem,
+ int *vq_num);
+extern VirtIOS390Device *s390_virtio_bus_find_mem(VirtIOS390Bus *bus,
+ ram_addr_t mem);
+
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* [Qemu-devel] [PATCH 08/11] Add S390x virtio machine description
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (6 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 07/11] Add S390x virtio machine bus Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-30 18:18 ` Aurelien Jarno
2009-11-26 13:23 ` [Qemu-devel] [PATCH 09/11] S390 GDB stub Alexander Graf
` (3 subsequent siblings)
11 siblings, 1 reply; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
In order to use the new S390x virtio bus we just introduced, we also
need a machine description that sets up the machine according to our
PV specification.
Let's add that machine description and be happy!
Signed-off-by: Alexander Graf <agraf@suse.de>
---
v4 -> v5
- disable early console
---
Makefile.target | 2 +-
hw/s390-virtio.c | 240 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 241 insertions(+), 1 deletions(-)
create mode 100644 hw/s390-virtio.c
diff --git a/Makefile.target b/Makefile.target
index 1619c23..9460c56 100644
--- a/Makefile.target
+++ b/Makefile.target
@@ -296,7 +296,7 @@ obj-sh4-y += ide/core.o ide/mmio.o
obj-m68k-y = an5206.o mcf5206.o mcf_uart.o mcf_intc.o mcf5208.o mcf_fec.o
obj-m68k-y += m68k-semi.o dummy_m68k.o
-obj-s390x-y = s390-virtio-bus.o
+obj-s390x-y = s390-virtio-bus.o s390-virtio.o
main.o vl.o: QEMU_CFLAGS+=$(GPROF_CFLAGS)
diff --git a/hw/s390-virtio.c b/hw/s390-virtio.c
new file mode 100644
index 0000000..a501e34
--- /dev/null
+++ b/hw/s390-virtio.c
@@ -0,0 +1,240 @@
+/*
+ * QEMU S390 virtio target
+ *
+ * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
+ *
+ * This library is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2 of the License, or (at your option) any later version.
+ *
+ * This library is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this library; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include "hw.h"
+#include "block.h"
+#include "sysemu.h"
+#include "net.h"
+#include "boards.h"
+#include "monitor.h"
+#include "loader.h"
+#include "elf.h"
+#include "hw/virtio.h"
+#include "hw/virtio-console.h"
+#include "hw/sysbus.h"
+#include "kvm.h"
+
+#include "hw/s390-virtio-bus.h"
+
+//#define DEBUG_S390
+
+#ifdef DEBUG_S390
+#define dprintf(fmt, ...) \
+ do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
+#else
+#define dprintf(fmt, ...) \
+ do { } while (0)
+#endif
+
+#define KVM_S390_VIRTIO_NOTIFY 0
+#define KVM_S390_VIRTIO_RESET 1
+#define KVM_S390_VIRTIO_SET_STATUS 2
+
+#define KERN_IMAGE_START 0x010000UL
+#define KERN_PARM_AREA 0x010480UL
+#define INITRD_START 0x800000UL
+#define INITRD_PARM_START 0x010408UL
+#define INITRD_PARM_SIZE 0x010410UL
+#define PARMFILE_START 0x001000UL
+
+#define MAX_BLK_DEVS 10
+
+static VirtIOS390Bus *s390_bus;
+static CPUState **ipi_states;
+
+void irq_info(Monitor *mon);
+void pic_info(Monitor *mon);
+
+void irq_info(Monitor *mon)
+{
+}
+
+void pic_info(Monitor *mon)
+{
+}
+
+CPUState *s390_cpu_addr2state(uint16_t cpu_addr)
+{
+ if (cpu_addr >= smp_cpus)
+ return NULL;
+
+ return ipi_states[cpu_addr];
+}
+
+int s390_virtio_hypercall(CPUState *env)
+{
+ int r = 0, i;
+ target_ulong mem = env->regs[2];
+
+ dprintf("KVM hypercall: %ld\n", env->regs[1]);
+ switch (env->regs[1]) {
+ case KVM_S390_VIRTIO_NOTIFY:
+ if (mem > ram_size) {
+ VirtIOS390Device *dev = s390_virtio_bus_find_vring(s390_bus,
+ mem, &i);
+ if (dev) {
+ virtio_queue_notify(dev->vdev, i);
+ } else {
+ r = -EINVAL;
+ }
+ } else {
+ /* Early printk */
+ }
+ break;
+ case KVM_S390_VIRTIO_RESET:
+ {
+ /* Virtio_reset resets the internal addresses, so we'd have to sync
+ them up again. We don't want to reallicate a vring though, so let's
+ just not reset. */
+ /* virtio_reset(dev->vdev); */
+ break;
+ }
+ case KVM_S390_VIRTIO_SET_STATUS:
+ {
+ VirtIOS390Device *dev;
+
+ dev = s390_virtio_bus_find_mem(s390_bus, mem);
+ if (dev)
+ s390_virtio_device_update_status(dev);
+ else
+ r = -EINVAL;
+ break;
+ }
+ default:
+ r = -EINVAL;
+ break;
+ }
+
+ env->regs[2] = r;
+ return 0;
+}
+
+/* PC hardware initialisation */
+static void s390_init(ram_addr_t _ram_size,
+ const char *boot_device,
+ const char *kernel_filename,
+ const char *kernel_cmdline,
+ const char *initrd_filename,
+ const char *cpu_model)
+{
+ CPUState *env = NULL;
+ ram_addr_t ram_addr, ram_size = _ram_size;
+ ram_addr_t kernel_size = 0;
+ ram_addr_t initrd_offset;
+ ram_addr_t initrd_size = 0;
+ int i;
+
+ /* get a BUS */
+ s390_bus = s390_virtio_bus_init(&ram_size);
+
+ /* allocate RAM */
+ ram_addr = qemu_ram_alloc(ram_size);
+ cpu_register_physical_memory(0, ram_size, ram_addr);
+
+ /* init CPUs */
+ if (cpu_model == NULL) {
+ cpu_model = "host";
+ }
+
+ ipi_states = qemu_malloc(sizeof(CPUState *) * smp_cpus);
+
+ for (i = 0; i < smp_cpus; i++) {
+ env = cpu_init(cpu_model);
+ ipi_states[smp_cpus - (i + 1)] = env;
+ env->halted = 1;
+ env->exception_index = EXCP_HLT;
+ }
+
+ env->halted = 0;
+ env->exception_index = 0;
+
+ if (kernel_filename) {
+ kernel_size = load_image(kernel_filename, qemu_get_ram_ptr(0));
+
+ if (lduw_phys(KERN_IMAGE_START) != 0x0dd0) {
+ fprintf(stderr, "Specified image is not an s390 boot image\n");
+ exit(1);
+ }
+
+ cpu_synchronize_state(env);
+ env->psw.addr = KERN_IMAGE_START;
+ env->psw.mask = 0x0000000180000000UL;
+ }
+
+ if (initrd_filename) {
+ initrd_offset = INITRD_START;
+ while (kernel_size + 0x100000 > initrd_offset)
+ initrd_offset += 0x100000;
+ initrd_size = load_image(initrd_filename, qemu_get_ram_ptr(initrd_offset));
+
+ stq_phys(INITRD_PARM_START, initrd_offset);
+ stq_phys(INITRD_PARM_SIZE, initrd_size);
+ }
+
+ if (kernel_cmdline) {
+ cpu_physical_memory_rw(KERN_PARM_AREA, (uint8_t *)kernel_cmdline,
+ strlen(kernel_cmdline), 1);
+ }
+
+ /* Create VirtIO console */
+ qdev_init_nofail(qdev_create((BusState *)s390_bus, "virtio-console-s390"));
+
+ /* Create VirtIO network adapters */
+ for(i = 0; i < nb_nics; i++) {
+ NICInfo *nd = &nd_table[i];
+ DeviceState *dev;
+
+ if (!nd->model)
+ nd->model = (char*)"virtio";
+
+ dev = qdev_create((BusState *)s390_bus, "virtio-net-s390");
+ qdev_set_nic_properties(dev, nd);
+ qdev_init_nofail(dev);
+ }
+
+ /* Create VirtIO disk drives */
+ for(i = 0; i < MAX_BLK_DEVS; i++) {
+ DriveInfo *dinfo;
+ DeviceState *dev;
+
+ dinfo = drive_get(IF_IDE, 0, i);
+ if (!dinfo)
+ continue;
+
+ dev = qdev_create((BusState *)s390_bus, "virtio-blk-s390");
+ qdev_prop_set_drive(dev, "drive", dinfo);
+ qdev_init_nofail(dev);
+ }
+}
+
+static QEMUMachine s390_machine = {
+ .name = "s390-virtio",
+ .alias = "s390",
+ .desc = "VirtIO based S390 machine",
+ .init = s390_init,
+ .max_cpus = 255,
+ .is_default = 1,
+};
+
+static void s390_machine_init(void)
+{
+ qemu_register_machine(&s390_machine);
+}
+
+machine_init(s390_machine_init);
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 08/11] Add S390x virtio machine description
2009-11-26 13:23 ` [Qemu-devel] [PATCH 08/11] Add S390x virtio machine description Alexander Graf
@ 2009-11-30 18:18 ` Aurelien Jarno
0 siblings, 0 replies; 49+ messages in thread
From: Aurelien Jarno @ 2009-11-30 18:18 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
On Thu, Nov 26, 2009 at 02:23:17PM +0100, Alexander Graf wrote:
> In order to use the new S390x virtio bus we just introduced, we also
> need a machine description that sets up the machine according to our
> PV specification.
>
> Let's add that machine description and be happy!
Please find the comments below.
> Signed-off-by: Alexander Graf <agraf@suse.de>
>
> ---
>
> v4 -> v5
>
> - disable early console
> ---
> Makefile.target | 2 +-
> hw/s390-virtio.c | 240 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
> 2 files changed, 241 insertions(+), 1 deletions(-)
> create mode 100644 hw/s390-virtio.c
>
> diff --git a/Makefile.target b/Makefile.target
> index 1619c23..9460c56 100644
> --- a/Makefile.target
> +++ b/Makefile.target
> @@ -296,7 +296,7 @@ obj-sh4-y += ide/core.o ide/mmio.o
> obj-m68k-y = an5206.o mcf5206.o mcf_uart.o mcf_intc.o mcf5208.o mcf_fec.o
> obj-m68k-y += m68k-semi.o dummy_m68k.o
>
> -obj-s390x-y = s390-virtio-bus.o
> +obj-s390x-y = s390-virtio-bus.o s390-virtio.o
>
> main.o vl.o: QEMU_CFLAGS+=$(GPROF_CFLAGS)
>
> diff --git a/hw/s390-virtio.c b/hw/s390-virtio.c
> new file mode 100644
> index 0000000..a501e34
> --- /dev/null
> +++ b/hw/s390-virtio.c
> @@ -0,0 +1,240 @@
> +/*
> + * QEMU S390 virtio target
> + *
> + * Copyright (c) 2009 Alexander Graf <agraf@suse.de>
> + *
> + * This library is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU Lesser General Public
> + * License as published by the Free Software Foundation; either
> + * version 2 of the License, or (at your option) any later version.
> + *
> + * This library is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
> + * Lesser General Public License for more details.
> + *
> + * You should have received a copy of the GNU Lesser General Public
> + * License along with this library; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include "hw.h"
> +#include "block.h"
> +#include "sysemu.h"
> +#include "net.h"
> +#include "boards.h"
> +#include "monitor.h"
> +#include "loader.h"
> +#include "elf.h"
> +#include "hw/virtio.h"
> +#include "hw/virtio-console.h"
> +#include "hw/sysbus.h"
> +#include "kvm.h"
> +
> +#include "hw/s390-virtio-bus.h"
> +
> +//#define DEBUG_S390
> +
> +#ifdef DEBUG_S390
> +#define dprintf(fmt, ...) \
> + do { fprintf(stderr, fmt, ## __VA_ARGS__); } while (0)
> +#else
> +#define dprintf(fmt, ...) \
> + do { } while (0)
> +#endif
> +
> +#define KVM_S390_VIRTIO_NOTIFY 0
> +#define KVM_S390_VIRTIO_RESET 1
> +#define KVM_S390_VIRTIO_SET_STATUS 2
> +
> +#define KERN_IMAGE_START 0x010000UL
> +#define KERN_PARM_AREA 0x010480UL
> +#define INITRD_START 0x800000UL
> +#define INITRD_PARM_START 0x010408UL
> +#define INITRD_PARM_SIZE 0x010410UL
> +#define PARMFILE_START 0x001000UL
> +
> +#define MAX_BLK_DEVS 10
> +
> +static VirtIOS390Bus *s390_bus;
> +static CPUState **ipi_states;
> +
> +void irq_info(Monitor *mon);
> +void pic_info(Monitor *mon);
> +
> +void irq_info(Monitor *mon)
> +{
> +}
> +
> +void pic_info(Monitor *mon)
> +{
> +}
> +
> +CPUState *s390_cpu_addr2state(uint16_t cpu_addr)
> +{
> + if (cpu_addr >= smp_cpus)
> + return NULL;
coding style.
> +
> + return ipi_states[cpu_addr];
> +}
> +
> +int s390_virtio_hypercall(CPUState *env)
> +{
> + int r = 0, i;
> + target_ulong mem = env->regs[2];
> +
> + dprintf("KVM hypercall: %ld\n", env->regs[1]);
> + switch (env->regs[1]) {
> + case KVM_S390_VIRTIO_NOTIFY:
> + if (mem > ram_size) {
> + VirtIOS390Device *dev = s390_virtio_bus_find_vring(s390_bus,
> + mem, &i);
> + if (dev) {
> + virtio_queue_notify(dev->vdev, i);
> + } else {
> + r = -EINVAL;
> + }
> + } else {
> + /* Early printk */
> + }
> + break;
> + case KVM_S390_VIRTIO_RESET:
> + {
> + /* Virtio_reset resets the internal addresses, so we'd have to sync
> + them up again. We don't want to reallicate a vring though, so let's
^^^^^ typo
> + just not reset. */
> + /* virtio_reset(dev->vdev); */
> + break;
> + }
> + case KVM_S390_VIRTIO_SET_STATUS:
> + {
> + VirtIOS390Device *dev;
> +
> + dev = s390_virtio_bus_find_mem(s390_bus, mem);
> + if (dev)
> + s390_virtio_device_update_status(dev);
> + else
> + r = -EINVAL;
coding style.
> + break;
> + }
> + default:
> + r = -EINVAL;
> + break;
> + }
> +
> + env->regs[2] = r;
> + return 0;
> +}
> +
> +/* PC hardware initialisation */
> +static void s390_init(ram_addr_t _ram_size,
> + const char *boot_device,
> + const char *kernel_filename,
> + const char *kernel_cmdline,
> + const char *initrd_filename,
> + const char *cpu_model)
> +{
> + CPUState *env = NULL;
> + ram_addr_t ram_addr, ram_size = _ram_size;
> + ram_addr_t kernel_size = 0;
> + ram_addr_t initrd_offset;
> + ram_addr_t initrd_size = 0;
> + int i;
> +
> + /* get a BUS */
> + s390_bus = s390_virtio_bus_init(&ram_size);
> +
> + /* allocate RAM */
> + ram_addr = qemu_ram_alloc(ram_size);
> + cpu_register_physical_memory(0, ram_size, ram_addr);
> +
> + /* init CPUs */
> + if (cpu_model == NULL) {
> + cpu_model = "host";
> + }
> +
> + ipi_states = qemu_malloc(sizeof(CPUState *) * smp_cpus);
> +
> + for (i = 0; i < smp_cpus; i++) {
> + env = cpu_init(cpu_model);
> + ipi_states[smp_cpus - (i + 1)] = env;
> + env->halted = 1;
> + env->exception_index = EXCP_HLT;
> + }
> +
> + env->halted = 0;
> + env->exception_index = 0;
> +
> + if (kernel_filename) {
> + kernel_size = load_image(kernel_filename, qemu_get_ram_ptr(0));
> +
You should probably use load_image_targphys() instead to reload the
kernel on reset.
> + if (lduw_phys(KERN_IMAGE_START) != 0x0dd0) {
> + fprintf(stderr, "Specified image is not an s390 boot image\n");
> + exit(1);
> + }
> +
> + cpu_synchronize_state(env);
> + env->psw.addr = KERN_IMAGE_START;
> + env->psw.mask = 0x0000000180000000UL;
Shouldn't it be done by registering a reset handler, so that this value
are restored on reset?
> + }
> +
> + if (initrd_filename) {
> + initrd_offset = INITRD_START;
> + while (kernel_size + 0x100000 > initrd_offset)
> + initrd_offset += 0x100000;
coding style.
> + initrd_size = load_image(initrd_filename, qemu_get_ram_ptr(initrd_offset));
> +
Same as for the kernel.
> + stq_phys(INITRD_PARM_START, initrd_offset);
> + stq_phys(INITRD_PARM_SIZE, initrd_size);
The same way, it should probably be added with rom_add_blob_fixed().
> + }
> +
> + if (kernel_cmdline) {
> + cpu_physical_memory_rw(KERN_PARM_AREA, (uint8_t *)kernel_cmdline,
> + strlen(kernel_cmdline), 1);
and here with pstrcpy_targphys().
> + }
> +
> + /* Create VirtIO console */
> + qdev_init_nofail(qdev_create((BusState *)s390_bus, "virtio-console-s390"));
> +
> + /* Create VirtIO network adapters */
> + for(i = 0; i < nb_nics; i++) {
> + NICInfo *nd = &nd_table[i];
> + DeviceState *dev;
> +
> + if (!nd->model)
> + nd->model = (char*)"virtio";
What happens if someone request a nice that is not virtio? It should
probably fail.
> + dev = qdev_create((BusState *)s390_bus, "virtio-net-s390");
> + qdev_set_nic_properties(dev, nd);
> + qdev_init_nofail(dev);
> + }
> +
> + /* Create VirtIO disk drives */
> + for(i = 0; i < MAX_BLK_DEVS; i++) {
> + DriveInfo *dinfo;
> + DeviceState *dev;
> +
> + dinfo = drive_get(IF_IDE, 0, i);
> + if (!dinfo)
> + continue;
> +
> + dev = qdev_create((BusState *)s390_bus, "virtio-blk-s390");
> + qdev_prop_set_drive(dev, "drive", dinfo);
> + qdev_init_nofail(dev);
> + }
> +}
> +
> +static QEMUMachine s390_machine = {
> + .name = "s390-virtio",
> + .alias = "s390",
> + .desc = "VirtIO based S390 machine",
> + .init = s390_init,
> + .max_cpus = 255,
> + .is_default = 1,
> +};
> +
> +static void s390_machine_init(void)
> +{
> + qemu_register_machine(&s390_machine);
> +}
> +
> +machine_init(s390_machine_init);
> --
> 1.6.0.2
>
>
>
>
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread
* [Qemu-devel] [PATCH 09/11] S390 GDB stub
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (7 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 08/11] Add S390x virtio machine description Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-26 13:23 ` [Qemu-devel] [PATCH 10/11] Set default console to virtio on S390x Alexander Graf
` (2 subsequent siblings)
11 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
In order to debug funny kernel breakages it's always good to have a working
gdb stub around.
While Uli's patches don't include one one, I needed one that's at least good
enough for 'bt' and some variable examinations during early bootup.
So here it is - the absolute basics to get the qemu gdb stub running with s390x
targets.
Sgined-off-by: Alexander Graf <agraf@suse.de>
---
gdbstub.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++++++++
1 files changed, 52 insertions(+), 0 deletions(-)
diff --git a/gdbstub.c b/gdbstub.c
index 055093f..1ba6ad0 100644
--- a/gdbstub.c
+++ b/gdbstub.c
@@ -1348,6 +1348,55 @@ static int cpu_gdb_write_register(CPUState *env, uint8_t *mem_buf, int n)
return 8;
}
+#elif defined (TARGET_S390X)
+
+#define NUM_CORE_REGS S390_NUM_TOTAL_REGS
+
+static int cpu_gdb_read_register(CPUState *env, uint8_t *mem_buf, int n)
+{
+ switch (n) {
+ case S390_PSWM_REGNUM: GET_REGL(env->psw.mask); break;
+ case S390_PSWA_REGNUM: GET_REGL(env->psw.addr); break;
+ case S390_R0_REGNUM ... S390_R15_REGNUM:
+ GET_REGL(env->regs[n-S390_R0_REGNUM]); break;
+ case S390_A0_REGNUM ... S390_A15_REGNUM:
+ GET_REG32(env->aregs[n-S390_A0_REGNUM]); break;
+ case S390_FPC_REGNUM: GET_REG32(env->fpc); break;
+ case S390_F0_REGNUM ... S390_F15_REGNUM:
+ /* XXX */
+ break;
+ case S390_PC_REGNUM: GET_REGL(env->psw.addr); break;
+ case S390_CC_REGNUM: GET_REG32(env->cc); break;
+ }
+
+ return 0;
+}
+
+static int cpu_gdb_write_register(CPUState *env, uint8_t *mem_buf, int n)
+{
+ target_ulong tmpl;
+ uint32_t tmp32;
+ int r = 8;
+ tmpl = ldtul_p(mem_buf);
+ tmp32 = ldl_p(mem_buf);
+
+ switch (n) {
+ case S390_PSWM_REGNUM: env->psw.mask = tmpl; break;
+ case S390_PSWA_REGNUM: env->psw.addr = tmpl; break;
+ case S390_R0_REGNUM ... S390_R15_REGNUM:
+ env->regs[n-S390_R0_REGNUM] = tmpl; break;
+ case S390_A0_REGNUM ... S390_A15_REGNUM:
+ env->aregs[n-S390_A0_REGNUM] = tmp32; r=4; break;
+ case S390_FPC_REGNUM: env->fpc = tmp32; r=4; break;
+ case S390_F0_REGNUM ... S390_F15_REGNUM:
+ /* XXX */
+ break;
+ case S390_PC_REGNUM: env->psw.addr = tmpl; break;
+ case S390_CC_REGNUM: env->cc = tmp32; r=4; break;
+ }
+
+ return r;
+}
#else
#define NUM_CORE_REGS 0
@@ -1616,6 +1665,9 @@ static void gdb_set_cpu_pc(GDBState *s, target_ulong pc)
s->c_cpu->pc = pc;
#elif defined (TARGET_ALPHA)
s->c_cpu->pc = pc;
+#elif defined (TARGET_S390X)
+ cpu_synchronize_state(s->c_cpu);
+ s->c_cpu->psw.addr = pc;
#endif
}
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* [Qemu-devel] [PATCH 10/11] Set default console to virtio on S390x
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (8 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 09/11] S390 GDB stub Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-26 13:23 ` [Qemu-devel] [PATCH 11/11] Add S390 maintainer information Alexander Graf
2009-11-30 18:18 ` [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Aurelien Jarno
11 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
All "normal" system emulation targets in qemu I'm aware of display output
on either VGA or serial output.
Our S390x virtio machine doesn't have such kind of legacy hardware. So
instead we need to default to a virtio console.
I'm not particularly proud of this patch. It would be a lot better to
have something in the machine description that tells us about the default
terminal.
Signed-off-by: Alexander Graf <agraf@suse.de>
---
vl.c | 26 ++++++++++++++++++++++++++
1 files changed, 26 insertions(+), 0 deletions(-)
diff --git a/vl.c b/vl.c
index ee43808..8703910 100644
--- a/vl.c
+++ b/vl.c
@@ -4663,6 +4663,20 @@ int main(int argc, char **argv, char **envp)
cyls = heads = secs = 0;
translation = BIOS_ATA_TRANSLATION_AUTO;
+#ifdef TARGET_S390X
+ for(i = 0; i < MAX_SERIAL_PORTS; i++)
+ serial_devices[i] = NULL;
+ serial_device_index = 0;
+
+ for(i = 0; i < MAX_PARALLEL_PORTS; i++)
+ parallel_devices[i] = NULL;
+ parallel_device_index = 0;
+
+ virtio_consoles[0] = "mon:stdio";
+ for(i = 1; i < MAX_VIRTIO_CONSOLES; i++)
+ virtio_consoles[i] = NULL;
+ virtio_console_index = 0;
+#else
serial_devices[0] = "vc:80Cx24C";
for(i = 1; i < MAX_SERIAL_PORTS; i++)
serial_devices[i] = NULL;
@@ -4676,6 +4690,7 @@ int main(int argc, char **argv, char **envp)
for(i = 0; i < MAX_VIRTIO_CONSOLES; i++)
virtio_consoles[i] = NULL;
virtio_console_index = 0;
+#endif
monitor_devices[0] = "vc:80Cx24C";
for (i = 1; i < MAX_MONITOR_DEVICES; i++) {
@@ -5585,6 +5600,17 @@ int main(int argc, char **argv, char **envp)
break;
}
}
+ for (i = 0; i < MAX_VIRTIO_CONSOLES; i++) {
+ const char *devname = virtio_consoles[i];
+ if (devname && !strcmp(devname,"mon:stdio")) {
+ monitor_devices[0] = NULL;
+ break;
+ } else if (devname && !strcmp(devname,"stdio")) {
+ monitor_devices[0] = NULL;
+ virtio_consoles[i] = "mon:stdio";
+ break;
+ }
+ }
}
if (nb_numa_nodes > 0) {
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* [Qemu-devel] [PATCH 11/11] Add S390 maintainer information
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (9 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 10/11] Set default console to virtio on S390x Alexander Graf
@ 2009-11-26 13:23 ` Alexander Graf
2009-11-30 18:18 ` [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Aurelien Jarno
11 siblings, 0 replies; 49+ messages in thread
From: Alexander Graf @ 2009-11-26 13:23 UTC (permalink / raw)
To: qemu-devel; +Cc: Carsten Otte, Aurelien Jarno
This patch adds information about who handles what when it comes to S390.
I'll gladly support anything that's related to the device emulation model and
S390 KVM parts.
Since this patchset doesn't implement S390 CPU emulation, I left that part
with a question mark. As soon as Uli's patchset gets committed I'd recommend
setting him there.
Signed-off-by: Alexander Graf <agraf@suse.de>
---
MAINTAINERS | 3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index d356495..56f107e 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -20,6 +20,7 @@ SH4 ?
CRIS Edgar E. Iglesias
Alpha ?
MicroBlaze Edgar E. Iglesias
+S390 ?
Machines (sorted by CPU):
-------------------------
@@ -63,6 +64,8 @@ CRIS
Alpha
MicroBlaze
petalogix_s3adsp1800.c Edgar E. Iglesias
+S390
+ s390-*.c Alexander Graf
Generic Subsystems:
-------------------
--
1.6.0.2
^ permalink raw reply related [flat|nested] 49+ messages in thread* Re: [Qemu-devel] [PATCH 00/11] S390x KVM support v4
2009-11-26 13:23 [Qemu-devel] [PATCH 00/11] S390x KVM support v4 Alexander Graf
` (10 preceding siblings ...)
2009-11-26 13:23 ` [Qemu-devel] [PATCH 11/11] Add S390 maintainer information Alexander Graf
@ 2009-11-30 18:18 ` Aurelien Jarno
11 siblings, 0 replies; 49+ messages in thread
From: Aurelien Jarno @ 2009-11-30 18:18 UTC (permalink / raw)
To: Alexander Graf; +Cc: Carsten Otte, qemu-devel
Hi,
On Thu, Nov 26, 2009 at 02:23:09PM +0100, Alexander Graf wrote:
> While S390x was one of the first targets that were supported by KVM it always
> lacked qemu system emulation support.
>
> In order to change that sad fact, I figured I'd just take on the task myself,
> taking kuli (http://www.ibm.com/developerworks/linux/linux390/kuli.html),
> Documentation/s390/kvm.txt and lguest as starting points to write a clean,
> qemu'ish S390x Virtio machine emulation.
>
> So now you can run Linux VMs on Linux on z/VM on LPAR on zSeries!
>
> This is the resulting code. Please comment on things you like and also on the
> ones you don't :-).
>
> The patchset is based on Uli's S390x userspace emulation patches. There's not
> really that much shared functionality, but I didn't want to reimplement the
> configure wheels. So make sure to have his patches applied first.
>
> Please only run S390x system virtualization using io threads. S390x has very
> few intercepts and even less that actually reach userspace. So you'll end up
> with a non-responsive VM without a dedicated IO thread.
>
> Use: $ ./configure --target-list=s390x-softmmu --enable-io-thread
>
> Currently only -kernel, -initrd and friends work. Booting from a real dasd
> device is not supported.
>
> Also to actually run this code you need a patch for an ugly bug in the kernel
> module that Carsten sent a patch to on the ML.
>
> I verified that the userspace actually works on a 2.6.27 (SLES11) kernel, so
> if it doesn't work on current git, please tell me! I'm too afraid that I might
> end up in a 3270 terminal to update the kernel on my z/VM instance :-).
>
> ##### WARNING #####
>
> Apparently the qemu review process has scared off yet another developer for now.
>
> So I'm sending this patchset without relying on Uli's set but rather
> implemented small stubs for TCG host and TCG guest support, so we can run
> on KVM. He promised me to send it again "soon" - whenever that is.
>
> These stubs don't serve any purpose but to get KVM running, so you get no x86
> on S390X or S390X on x86 for now!
>
This patchset basically looks ok. I have some minor comments, a lot of
them being related to coding style, but they should be easy to solve,
and should not prevent inclusion in 0.12.0 release.
Please find comments in the individual patches.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
^ permalink raw reply [flat|nested] 49+ messages in thread