* [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2
@ 2009-05-12 19:26 Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 1/3] target-ppc: expose cpu capability flags Nathan Froyd
` (3 more replies)
0 siblings, 4 replies; 10+ messages in thread
From: Nathan Froyd @ 2009-05-12 19:26 UTC (permalink / raw)
To: qemu-devel
This patch series adds signal handling support for 32-bit PPC linux-user
emulation. It requires a bit of shuffling of target-ppc code to ensure
that we know what the capabilities of our current CPU are so we know
which registers to save.
The only change from the previous version, posted here:
http://lists.gnu.org/archive/html/qemu-devel/2009-04/index.html
is that the PT_* constants have been renamed to TARGET_PT_* constants so
that everything will build OK on PowerPC Linux hosts.
The signal handling code has been tested by an in-house NPTL
implementation for PPC linux-user emulation. The NPTL implementation
will be coming along shortly, but the signal handling bits are useful on
their own.
-Nathan
^ permalink raw reply [flat|nested] 10+ messages in thread
* [Qemu-devel] [PATCH 1/3] target-ppc: expose cpu capability flags
2009-05-12 19:26 [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 Nathan Froyd
@ 2009-05-12 19:26 ` Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 2/3] linux-user: ppc signal handling Nathan Froyd
` (2 subsequent siblings)
3 siblings, 0 replies; 10+ messages in thread
From: Nathan Froyd @ 2009-05-12 19:26 UTC (permalink / raw)
To: qemu-devel
Do this so other pieces of code can make decisions based on the
capabilities of the CPU we're emulating.
Signed-off-by: Nathan Froyd <froydnj@codesourcery.com>
---
target-ppc/cpu.h | 139 +++++++++++++++++++++++++++++++++++++++++++
target-ppc/translate.c | 138 ------------------------------------------
target-ppc/translate_init.c | 1 +
3 files changed, 140 insertions(+), 138 deletions(-)
diff --git a/target-ppc/cpu.h b/target-ppc/cpu.h
index 0e8b49f..8dbc45d 100644
--- a/target-ppc/cpu.h
+++ b/target-ppc/cpu.h
@@ -642,6 +642,7 @@ struct CPUPPCState {
powerpc_input_t bus_model;
int bfd_mach;
uint32_t flags;
+ uint64_t insns_flags;
int error_code;
uint32_t pending_interrupts;
@@ -1320,6 +1321,144 @@ static inline void cpu_clone_regs(CPUState *env, target_ulong newsp)
#define SPR_E500_SVR (0x3FF)
/*****************************************************************************/
+/* PowerPC Instructions types definitions */
+enum {
+ PPC_NONE = 0x0000000000000000ULL,
+ /* PowerPC base instructions set */
+ PPC_INSNS_BASE = 0x0000000000000001ULL,
+ /* integer operations instructions */
+#define PPC_INTEGER PPC_INSNS_BASE
+ /* flow control instructions */
+#define PPC_FLOW PPC_INSNS_BASE
+ /* virtual memory instructions */
+#define PPC_MEM PPC_INSNS_BASE
+ /* ld/st with reservation instructions */
+#define PPC_RES PPC_INSNS_BASE
+ /* spr/msr access instructions */
+#define PPC_MISC PPC_INSNS_BASE
+ /* Deprecated instruction sets */
+ /* Original POWER instruction set */
+ PPC_POWER = 0x0000000000000002ULL,
+ /* POWER2 instruction set extension */
+ PPC_POWER2 = 0x0000000000000004ULL,
+ /* Power RTC support */
+ PPC_POWER_RTC = 0x0000000000000008ULL,
+ /* Power-to-PowerPC bridge (601) */
+ PPC_POWER_BR = 0x0000000000000010ULL,
+ /* 64 bits PowerPC instruction set */
+ PPC_64B = 0x0000000000000020ULL,
+ /* New 64 bits extensions (PowerPC 2.0x) */
+ PPC_64BX = 0x0000000000000040ULL,
+ /* 64 bits hypervisor extensions */
+ PPC_64H = 0x0000000000000080ULL,
+ /* New wait instruction (PowerPC 2.0x) */
+ PPC_WAIT = 0x0000000000000100ULL,
+ /* Time base mftb instruction */
+ PPC_MFTB = 0x0000000000000200ULL,
+
+ /* Fixed-point unit extensions */
+ /* PowerPC 602 specific */
+ PPC_602_SPEC = 0x0000000000000400ULL,
+ /* isel instruction */
+ PPC_ISEL = 0x0000000000000800ULL,
+ /* popcntb instruction */
+ PPC_POPCNTB = 0x0000000000001000ULL,
+ /* string load / store */
+ PPC_STRING = 0x0000000000002000ULL,
+
+ /* Floating-point unit extensions */
+ /* Optional floating point instructions */
+ PPC_FLOAT = 0x0000000000010000ULL,
+ /* New floating-point extensions (PowerPC 2.0x) */
+ PPC_FLOAT_EXT = 0x0000000000020000ULL,
+ PPC_FLOAT_FSQRT = 0x0000000000040000ULL,
+ PPC_FLOAT_FRES = 0x0000000000080000ULL,
+ PPC_FLOAT_FRSQRTE = 0x0000000000100000ULL,
+ PPC_FLOAT_FRSQRTES = 0x0000000000200000ULL,
+ PPC_FLOAT_FSEL = 0x0000000000400000ULL,
+ PPC_FLOAT_STFIWX = 0x0000000000800000ULL,
+
+ /* Vector/SIMD extensions */
+ /* Altivec support */
+ PPC_ALTIVEC = 0x0000000001000000ULL,
+ /* PowerPC 2.03 SPE extension */
+ PPC_SPE = 0x0000000002000000ULL,
+ /* PowerPC 2.03 SPE single-precision floating-point extension */
+ PPC_SPE_SINGLE = 0x0000000004000000ULL,
+ /* PowerPC 2.03 SPE double-precision floating-point extension */
+ PPC_SPE_DOUBLE = 0x0000000008000000ULL,
+
+ /* Optional memory control instructions */
+ PPC_MEM_TLBIA = 0x0000000010000000ULL,
+ PPC_MEM_TLBIE = 0x0000000020000000ULL,
+ PPC_MEM_TLBSYNC = 0x0000000040000000ULL,
+ /* sync instruction */
+ PPC_MEM_SYNC = 0x0000000080000000ULL,
+ /* eieio instruction */
+ PPC_MEM_EIEIO = 0x0000000100000000ULL,
+
+ /* Cache control instructions */
+ PPC_CACHE = 0x0000000200000000ULL,
+ /* icbi instruction */
+ PPC_CACHE_ICBI = 0x0000000400000000ULL,
+ /* dcbz instruction with fixed cache line size */
+ PPC_CACHE_DCBZ = 0x0000000800000000ULL,
+ /* dcbz instruction with tunable cache line size */
+ PPC_CACHE_DCBZT = 0x0000001000000000ULL,
+ /* dcba instruction */
+ PPC_CACHE_DCBA = 0x0000002000000000ULL,
+ /* Freescale cache locking instructions */
+ PPC_CACHE_LOCK = 0x0000004000000000ULL,
+
+ /* MMU related extensions */
+ /* external control instructions */
+ PPC_EXTERN = 0x0000010000000000ULL,
+ /* segment register access instructions */
+ PPC_SEGMENT = 0x0000020000000000ULL,
+ /* PowerPC 6xx TLB management instructions */
+ PPC_6xx_TLB = 0x0000040000000000ULL,
+ /* PowerPC 74xx TLB management instructions */
+ PPC_74xx_TLB = 0x0000080000000000ULL,
+ /* PowerPC 40x TLB management instructions */
+ PPC_40x_TLB = 0x0000100000000000ULL,
+ /* segment register access instructions for PowerPC 64 "bridge" */
+ PPC_SEGMENT_64B = 0x0000200000000000ULL,
+ /* SLB management */
+ PPC_SLBI = 0x0000400000000000ULL,
+
+ /* Embedded PowerPC dedicated instructions */
+ PPC_WRTEE = 0x0001000000000000ULL,
+ /* PowerPC 40x exception model */
+ PPC_40x_EXCP = 0x0002000000000000ULL,
+ /* PowerPC 405 Mac instructions */
+ PPC_405_MAC = 0x0004000000000000ULL,
+ /* PowerPC 440 specific instructions */
+ PPC_440_SPEC = 0x0008000000000000ULL,
+ /* BookE (embedded) PowerPC specification */
+ PPC_BOOKE = 0x0010000000000000ULL,
+ /* mfapidi instruction */
+ PPC_MFAPIDI = 0x0020000000000000ULL,
+ /* tlbiva instruction */
+ PPC_TLBIVA = 0x0040000000000000ULL,
+ /* tlbivax instruction */
+ PPC_TLBIVAX = 0x0080000000000000ULL,
+ /* PowerPC 4xx dedicated instructions */
+ PPC_4xx_COMMON = 0x0100000000000000ULL,
+ /* PowerPC 40x ibct instructions */
+ PPC_40x_ICBT = 0x0200000000000000ULL,
+ /* rfmci is not implemented in all BookE PowerPC */
+ PPC_RFMCI = 0x0400000000000000ULL,
+ /* rfdi instruction */
+ PPC_RFDI = 0x0800000000000000ULL,
+ /* DCR accesses */
+ PPC_DCR = 0x1000000000000000ULL,
+ /* DCR extended accesse */
+ PPC_DCRX = 0x2000000000000000ULL,
+ /* user-mode DCR access, implemented in PowerPC 460 */
+ PPC_DCRUX = 0x4000000000000000ULL,
+};
+
+/*****************************************************************************/
/* Memory access type :
* may be needed for precise access rights control and precise exceptions.
*/
diff --git a/target-ppc/translate.c b/target-ppc/translate.c
index 24c78d1..2cb90f0 100644
--- a/target-ppc/translate.c
+++ b/target-ppc/translate.c
@@ -457,144 +457,6 @@ static always_inline target_ulong MASK (uint32_t start, uint32_t end)
}
/*****************************************************************************/
-/* PowerPC Instructions types definitions */
-enum {
- PPC_NONE = 0x0000000000000000ULL,
- /* PowerPC base instructions set */
- PPC_INSNS_BASE = 0x0000000000000001ULL,
- /* integer operations instructions */
-#define PPC_INTEGER PPC_INSNS_BASE
- /* flow control instructions */
-#define PPC_FLOW PPC_INSNS_BASE
- /* virtual memory instructions */
-#define PPC_MEM PPC_INSNS_BASE
- /* ld/st with reservation instructions */
-#define PPC_RES PPC_INSNS_BASE
- /* spr/msr access instructions */
-#define PPC_MISC PPC_INSNS_BASE
- /* Deprecated instruction sets */
- /* Original POWER instruction set */
- PPC_POWER = 0x0000000000000002ULL,
- /* POWER2 instruction set extension */
- PPC_POWER2 = 0x0000000000000004ULL,
- /* Power RTC support */
- PPC_POWER_RTC = 0x0000000000000008ULL,
- /* Power-to-PowerPC bridge (601) */
- PPC_POWER_BR = 0x0000000000000010ULL,
- /* 64 bits PowerPC instruction set */
- PPC_64B = 0x0000000000000020ULL,
- /* New 64 bits extensions (PowerPC 2.0x) */
- PPC_64BX = 0x0000000000000040ULL,
- /* 64 bits hypervisor extensions */
- PPC_64H = 0x0000000000000080ULL,
- /* New wait instruction (PowerPC 2.0x) */
- PPC_WAIT = 0x0000000000000100ULL,
- /* Time base mftb instruction */
- PPC_MFTB = 0x0000000000000200ULL,
-
- /* Fixed-point unit extensions */
- /* PowerPC 602 specific */
- PPC_602_SPEC = 0x0000000000000400ULL,
- /* isel instruction */
- PPC_ISEL = 0x0000000000000800ULL,
- /* popcntb instruction */
- PPC_POPCNTB = 0x0000000000001000ULL,
- /* string load / store */
- PPC_STRING = 0x0000000000002000ULL,
-
- /* Floating-point unit extensions */
- /* Optional floating point instructions */
- PPC_FLOAT = 0x0000000000010000ULL,
- /* New floating-point extensions (PowerPC 2.0x) */
- PPC_FLOAT_EXT = 0x0000000000020000ULL,
- PPC_FLOAT_FSQRT = 0x0000000000040000ULL,
- PPC_FLOAT_FRES = 0x0000000000080000ULL,
- PPC_FLOAT_FRSQRTE = 0x0000000000100000ULL,
- PPC_FLOAT_FRSQRTES = 0x0000000000200000ULL,
- PPC_FLOAT_FSEL = 0x0000000000400000ULL,
- PPC_FLOAT_STFIWX = 0x0000000000800000ULL,
-
- /* Vector/SIMD extensions */
- /* Altivec support */
- PPC_ALTIVEC = 0x0000000001000000ULL,
- /* PowerPC 2.03 SPE extension */
- PPC_SPE = 0x0000000002000000ULL,
- /* PowerPC 2.03 SPE single-precision floating-point extension */
- PPC_SPE_SINGLE = 0x0000000004000000ULL,
- /* PowerPC 2.03 SPE double-precision floating-point extension */
- PPC_SPE_DOUBLE = 0x0000000008000000ULL,
-
- /* Optional memory control instructions */
- PPC_MEM_TLBIA = 0x0000000010000000ULL,
- PPC_MEM_TLBIE = 0x0000000020000000ULL,
- PPC_MEM_TLBSYNC = 0x0000000040000000ULL,
- /* sync instruction */
- PPC_MEM_SYNC = 0x0000000080000000ULL,
- /* eieio instruction */
- PPC_MEM_EIEIO = 0x0000000100000000ULL,
-
- /* Cache control instructions */
- PPC_CACHE = 0x0000000200000000ULL,
- /* icbi instruction */
- PPC_CACHE_ICBI = 0x0000000400000000ULL,
- /* dcbz instruction with fixed cache line size */
- PPC_CACHE_DCBZ = 0x0000000800000000ULL,
- /* dcbz instruction with tunable cache line size */
- PPC_CACHE_DCBZT = 0x0000001000000000ULL,
- /* dcba instruction */
- PPC_CACHE_DCBA = 0x0000002000000000ULL,
- /* Freescale cache locking instructions */
- PPC_CACHE_LOCK = 0x0000004000000000ULL,
-
- /* MMU related extensions */
- /* external control instructions */
- PPC_EXTERN = 0x0000010000000000ULL,
- /* segment register access instructions */
- PPC_SEGMENT = 0x0000020000000000ULL,
- /* PowerPC 6xx TLB management instructions */
- PPC_6xx_TLB = 0x0000040000000000ULL,
- /* PowerPC 74xx TLB management instructions */
- PPC_74xx_TLB = 0x0000080000000000ULL,
- /* PowerPC 40x TLB management instructions */
- PPC_40x_TLB = 0x0000100000000000ULL,
- /* segment register access instructions for PowerPC 64 "bridge" */
- PPC_SEGMENT_64B = 0x0000200000000000ULL,
- /* SLB management */
- PPC_SLBI = 0x0000400000000000ULL,
-
- /* Embedded PowerPC dedicated instructions */
- PPC_WRTEE = 0x0001000000000000ULL,
- /* PowerPC 40x exception model */
- PPC_40x_EXCP = 0x0002000000000000ULL,
- /* PowerPC 405 Mac instructions */
- PPC_405_MAC = 0x0004000000000000ULL,
- /* PowerPC 440 specific instructions */
- PPC_440_SPEC = 0x0008000000000000ULL,
- /* BookE (embedded) PowerPC specification */
- PPC_BOOKE = 0x0010000000000000ULL,
- /* mfapidi instruction */
- PPC_MFAPIDI = 0x0020000000000000ULL,
- /* tlbiva instruction */
- PPC_TLBIVA = 0x0040000000000000ULL,
- /* tlbivax instruction */
- PPC_TLBIVAX = 0x0080000000000000ULL,
- /* PowerPC 4xx dedicated instructions */
- PPC_4xx_COMMON = 0x0100000000000000ULL,
- /* PowerPC 40x ibct instructions */
- PPC_40x_ICBT = 0x0200000000000000ULL,
- /* rfmci is not implemented in all BookE PowerPC */
- PPC_RFMCI = 0x0400000000000000ULL,
- /* rfdi instruction */
- PPC_RFDI = 0x0800000000000000ULL,
- /* DCR accesses */
- PPC_DCR = 0x1000000000000000ULL,
- /* DCR extended accesse */
- PPC_DCRX = 0x2000000000000000ULL,
- /* user-mode DCR access, implemented in PowerPC 460 */
- PPC_DCRUX = 0x4000000000000000ULL,
-};
-
-/*****************************************************************************/
/* PowerPC instructions table */
#if HOST_LONG_BITS == 64
#define OPC_ALIGN 8
diff --git a/target-ppc/translate_init.c b/target-ppc/translate_init.c
index f5e3b28..e73bed2 100644
--- a/target-ppc/translate_init.c
+++ b/target-ppc/translate_init.c
@@ -9483,6 +9483,7 @@ int cpu_ppc_register_internal (CPUPPCState *env, const ppc_def_t *def)
env->mmu_model = def->mmu_model;
env->excp_model = def->excp_model;
env->bus_model = def->bus_model;
+ env->insns_flags = def->insns_flags;
env->flags = def->flags;
env->bfd_mach = def->bfd_mach;
env->check_pow = def->check_pow;
--
1.6.0.5
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [Qemu-devel] [PATCH 2/3] linux-user: ppc signal handling
2009-05-12 19:26 [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 1/3] target-ppc: expose cpu capability flags Nathan Froyd
@ 2009-05-12 19:26 ` Nathan Froyd
2009-05-13 2:13 ` Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 3/3] support ELF_HWCAP for PPPC Nathan Froyd
2009-05-12 23:13 ` [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 malc
3 siblings, 1 reply; 10+ messages in thread
From: Nathan Froyd @ 2009-05-12 19:26 UTC (permalink / raw)
To: qemu-devel
Implement setup_{,rt_}frame and do_{,rt_}sigreturn for PPC 32-bit. Use
the same TARGET_QEMU_ESIGRETURN hack as for MIPS to avoid clobbering
register state on a sigreturn.
Signed-off-by: Nathan Froyd <froydnj@codesourcery.com>
---
linux-user/main.c | 5 +
linux-user/ppc/syscall.h | 3 +
linux-user/signal.c | 596 ++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 604 insertions(+), 0 deletions(-)
diff --git a/linux-user/main.c b/linux-user/main.c
index dc39b05..9c57c17 100644
--- a/linux-user/main.c
+++ b/linux-user/main.c
@@ -1461,6 +1461,11 @@ void cpu_loop(CPUPPCState *env)
ret = do_syscall(env, env->gpr[0], env->gpr[3], env->gpr[4],
env->gpr[5], env->gpr[6], env->gpr[7],
env->gpr[8]);
+ if (ret == (uint32_t)(-TARGET_QEMU_ESIGRETURN)) {
+ /* Returning from a successful sigreturn syscall.
+ Avoid corrupting register state. */
+ break;
+ }
if (ret > (uint32_t)(-515)) {
env->crf[0] |= 0x1;
ret = -ret;
diff --git a/linux-user/ppc/syscall.h b/linux-user/ppc/syscall.h
index a21d61b..d789911 100644
--- a/linux-user/ppc/syscall.h
+++ b/linux-user/ppc/syscall.h
@@ -51,6 +51,9 @@ struct target_revectored_struct {
abi_ulong __map[8]; /* 256 bits */
};
+/* Nasty hack: define a fake errno value for use by sigreturn. */
+#define TARGET_QEMU_ESIGRETURN 255
+
/*
* flags masks
*/
diff --git a/linux-user/signal.c b/linux-user/signal.c
index 742d52a..d7dd3cf 100644
--- a/linux-user/signal.c
+++ b/linux-user/signal.c
@@ -3182,6 +3182,602 @@ long do_rt_sigreturn(CPUState *env)
return -TARGET_ENOSYS;
}
+#elif defined(TARGET_PPC) && !defined(TARGET_PPC64)
+
+/* FIXME: Many of the structures are defined for both PPC and PPC64, but
+ the signal handling is different enough that we haven't implemented
+ support for PPC64 yet. Hence the restriction above.
+
+ There are various #if'd blocks for code for TARGET_PPC64. These
+ blocks should go away so that we can successfully run 32-bit and
+ 64-bit binaries on a QEMU configured for PPC64. */
+
+/* Size of dummy stack frame allocated when calling signal handler.
+ See arch/powerpc/include/asm/ptrace.h. */
+#if defined(TARGET_PPC64)
+#define SIGNAL_FRAMESIZE 128
+#else
+#define SIGNAL_FRAMESIZE 64
+#endif
+
+/* See arch/powerpc/include/asm/sigcontext.h. */
+struct target_sigcontext {
+ target_ulong _unused[4];
+ int32_t signal;
+#if defined(TARGET_PPC64)
+ int32_t pad0;
+#endif
+ target_ulong handler;
+ target_ulong oldmask;
+ target_ulong regs; /* struct pt_regs __user * */
+ /* TODO: PPC64 includes extra bits here. */
+};
+
+/* Indices for target_mcontext.mc_gregs, below.
+ See arch/powerpc/include/asm/ptrace.h for details. */
+enum {
+ TARGET_PT_R0 = 0,
+ TARGET_PT_R1 = 1,
+ TARGET_PT_R2 = 2,
+ TARGET_PT_R3 = 3,
+ TARGET_PT_R4 = 4,
+ TARGET_PT_R5 = 5,
+ TARGET_PT_R6 = 6,
+ TARGET_PT_R7 = 7,
+ TARGET_PT_R8 = 8,
+ TARGET_PT_R9 = 9,
+ TARGET_PT_R10 = 10,
+ TARGET_PT_R11 = 11,
+ TARGET_PT_R12 = 12,
+ TARGET_PT_R13 = 13,
+ TARGET_PT_R14 = 14,
+ TARGET_PT_R15 = 15,
+ TARGET_PT_R16 = 16,
+ TARGET_PT_R17 = 17,
+ TARGET_PT_R18 = 18,
+ TARGET_PT_R19 = 19,
+ TARGET_PT_R20 = 20,
+ TARGET_PT_R21 = 21,
+ TARGET_PT_R22 = 22,
+ TARGET_PT_R23 = 23,
+ TARGET_PT_R24 = 24,
+ TARGET_PT_R25 = 25,
+ TARGET_PT_R26 = 26,
+ TARGET_PT_R27 = 27,
+ TARGET_PT_R28 = 28,
+ TARGET_PT_R29 = 29,
+ TARGET_PT_R30 = 30,
+ TARGET_PT_R31 = 31,
+ TARGET_PT_NIP = 32,
+ TARGET_PT_MSR = 33,
+ TARGET_PT_ORIG_R3 = 34,
+ TARGET_PT_CTR = 35,
+ TARGET_PT_LNK = 36,
+ TARGET_PT_XER = 37,
+ TARGET_PT_CCR = 38,
+ /* Yes, there are two registers with #39. One is 64-bit only. */
+ TARGET_PT_MQ = 39,
+ TARGET_PT_SOFTE = 39,
+ TARGET_PT_TRAP = 40,
+ TARGET_PT_DAR = 41,
+ TARGET_PT_DSISR = 42,
+ TARGET_PT_RESULT = 43,
+ TARGET_PT_REGS_COUNT = 44
+};
+
+/* See arch/powerpc/include/asm/ucontext.h. Only used for 32-bit PPC;
+ on 64-bit PPC, sigcontext and mcontext are one and the same. */
+struct target_mcontext {
+ target_ulong mc_gregs[48];
+ /* Includes fpscr. */
+ uint64_t mc_fregs[33];
+ target_ulong mc_pad[2];
+ /* We need to handle Altivec and SPE at the same time, which no
+ kernel needs to do. Fortunately, the kernel defines this bit to
+ be Altivec-register-large all the time, rather than trying to
+ twiddle it based on the specific platform. */
+ union {
+ /* SPE vector registers. One extra for SPEFSCR. */
+ uint32_t spe[33];
+ /* Altivec vector registers. The packing of VSCR and VRSAVE
+ varies depending on whether we're PPC64 or not: PPC64 splits
+ them apart; PPC32 stuffs them together. */
+#if defined(TARGET_PPC64)
+#define NVRREG 34
+#else
+#define NVRREG 33
+#endif
+ ppc_avr_t altivec[NVRREG];
+#undef NVRREG
+ } mc_vregs __attribute__((__aligned__(16)));
+};
+
+struct target_ucontext {
+ target_ulong uc_flags;
+ target_ulong uc_link; /* struct ucontext __user * */
+ struct target_sigaltstack uc_stack;
+#if !defined(TARGET_PPC64)
+ int32_t uc_pad[7];
+ target_ulong uc_regs; /* struct mcontext __user *
+ points to uc_mcontext field */
+#endif
+ target_sigset_t uc_sigmask;
+#if defined(TARGET_PPC64)
+ target_sigset_t unused[15]; /* Allow for uc_sigmask growth */
+ struct target_sigcontext uc_mcontext;
+#else
+ int32_t uc_maskext[30];
+ int32_t uc_pad2[3];
+ struct target_mcontext uc_mcontext;
+#endif
+};
+
+/* See arch/powerpc/kernel/signal_32.c. */
+struct target_sigframe {
+ struct target_sigcontext sctx;
+ struct target_mcontext mctx;
+ int32_t abigap[56];
+};
+
+struct target_rt_sigframe {
+ struct target_siginfo info;
+ struct target_ucontext uc;
+ int32_t abigap[56];
+};
+
+/* We use the mc_pad field for the signal return trampoline. */
+#define tramp mc_pad
+
+/* See arch/powerpc/kernel/signal.c. */
+static target_ulong get_sigframe(struct target_sigaction *ka,
+ CPUState *env,
+ int frame_size)
+{
+ target_ulong oldsp, newsp;
+
+ oldsp = env->gpr[1];
+
+ if ((ka->sa_flags & TARGET_SA_ONSTACK) &&
+ (sas_ss_flags(oldsp))) {
+ oldsp = (target_sigaltstack_used.ss_sp
+ + target_sigaltstack_used.ss_size);
+ }
+
+ newsp = (oldsp - frame_size) & ~0xFUL;
+
+ return newsp;
+}
+
+static int save_user_regs(CPUState *env, struct target_mcontext *frame,
+ int sigret)
+{
+ target_ulong msr = env->msr;
+ int i;
+ target_ulong ccr = 0;
+
+ /* In general, the kernel attempts to be intelligent about what it
+ needs to save for Altivec/FP/SPE registers. We don't care that
+ much, so we just go ahead and save everything. */
+
+ /* Save general registers. */
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ if (__put_user(env->gpr[i], &frame->mc_gregs[i])) {
+ return 1;
+ }
+ }
+ if (__put_user(env->nip, &frame->mc_gregs[TARGET_PT_NIP])
+ || __put_user(env->ctr, &frame->mc_gregs[TARGET_PT_CTR])
+ || __put_user(env->lr, &frame->mc_gregs[TARGET_PT_LNK])
+ || __put_user(env->xer, &frame->mc_gregs[TARGET_PT_XER]))
+ return 1;
+
+ for (i = 0; i < ARRAY_SIZE(env->crf); i++) {
+ ccr |= env->crf[i] << (32 - ((i + 1) * 4));
+ }
+ if (__put_user(ccr, &frame->mc_gregs[TARGET_PT_CCR]))
+ return 1;
+
+ /* Save Altivec registers if necessary. */
+ if (env->insns_flags & PPC_ALTIVEC) {
+ for (i = 0; i < ARRAY_SIZE(env->avr); i++) {
+ ppc_avr_t *avr = &env->avr[i];
+ ppc_avr_t *vreg = &frame->mc_vregs.altivec[i];
+
+ if (__put_user(avr->u64[0], &vreg->u64[0]) ||
+ __put_user(avr->u64[1], &vreg->u64[1])) {
+ return 1;
+ }
+ }
+ /* Set MSR_VR in the saved MSR value to indicate that
+ frame->mc_vregs contains valid data. */
+ msr |= MSR_VR;
+ if (__put_user((uint32_t)env->spr[SPR_VRSAVE],
+ &frame->mc_vregs.altivec[32].u32[3]))
+ return 1;
+ }
+
+ /* Save floating point registers. */
+ if (env->insns_flags & PPC_FLOAT) {
+ for (i = 0; i < ARRAY_SIZE(env->fpr); i++) {
+ if (__put_user(env->fpr[i], &frame->mc_fregs[i])) {
+ return 1;
+ }
+ }
+ if (__put_user((uint64_t) env->fpscr, &frame->mc_fregs[32]))
+ return 1;
+ }
+
+ /* Save SPE registers. The kernel only saves the high half. */
+ if (env->insns_flags & PPC_SPE) {
+#if defined(TARGET_PPC64)
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ if (__put_user(env->gpr[i] >> 32, &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ }
+#else
+ for (i = 0; i < ARRAY_SIZE(env->gprh); i++) {
+ if (__put_user(env->gprh[i], &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ }
+#endif
+ /* Set MSR_SPE in the saved MSR value to indicate that
+ frame->mc_vregs contains valid data. */
+ msr |= MSR_SPE;
+ if (__put_user(env->spe_fscr, &frame->mc_vregs.spe[32]))
+ return 1;
+ }
+
+ /* Store MSR. */
+ if (__put_user(msr, &frame->mc_gregs[TARGET_PT_MSR]))
+ return 1;
+
+ /* Set up the sigreturn trampoline: li r0,sigret; sc. */
+ if (sigret) {
+ if (__put_user(0x38000000UL | sigret, &frame->tramp[0]) ||
+ __put_user(0x44000002UL, &frame->tramp[1])) {
+ return 1;
+ }
+ }
+
+ return 0;
+}
+
+static int restore_user_regs(CPUState *env,
+ struct target_mcontext *frame, int sig)
+{
+ target_ulong save_r2 = 0;
+ target_ulong msr;
+ target_ulong ccr;
+
+ int i;
+
+ if (!sig) {
+ save_r2 = env->gpr[2];
+ }
+
+ /* Restore general registers. */
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ if (__get_user(env->gpr[i], &frame->mc_gregs[i])) {
+ return 1;
+ }
+ }
+ if (__get_user(env->nip, &frame->mc_gregs[TARGET_PT_NIP])
+ || __get_user(env->ctr, &frame->mc_gregs[TARGET_PT_CTR])
+ || __get_user(env->lr, &frame->mc_gregs[TARGET_PT_LNK])
+ || __get_user(env->xer, &frame->mc_gregs[TARGET_PT_XER]))
+ return 1;
+ if (__get_user(ccr, &frame->mc_gregs[TARGET_PT_CCR]))
+ return 1;
+
+ for (i = 0; i < ARRAY_SIZE(env->crf); i++) {
+ env->crf[i] = (ccr >> (32 - ((i + 1) * 4))) & 0xf;
+ }
+
+ if (!sig) {
+ env->gpr[2] = save_r2;
+ }
+ /* Restore MSR. */
+ if (__get_user(msr, &frame->mc_gregs[TARGET_PT_MSR]))
+ return 1;
+
+ /* If doing signal return, restore the previous little-endian mode. */
+ if (sig)
+ env->msr = (env->msr & ~MSR_LE) | (msr & MSR_LE);
+
+ /* Restore Altivec registers if necessary. */
+ if (env->insns_flags & PPC_ALTIVEC) {
+ for (i = 0; i < ARRAY_SIZE(env->avr); i++) {
+ ppc_avr_t *avr = &env->avr[i];
+ ppc_avr_t *vreg = &frame->mc_vregs.altivec[i];
+
+ if (__get_user(avr->u64[0], &vreg->u64[0]) ||
+ __get_user(avr->u64[1], &vreg->u64[1])) {
+ return 1;
+ }
+ }
+ /* Set MSR_VEC in the saved MSR value to indicate that
+ frame->mc_vregs contains valid data. */
+ if (__get_user(env->spr[SPR_VRSAVE],
+ (target_ulong *)(&frame->mc_vregs.altivec[32].u32[3])))
+ return 1;
+ }
+
+ /* Restore floating point registers. */
+ if (env->insns_flags & PPC_FLOAT) {
+ uint64_t fpscr;
+ for (i = 0; i < ARRAY_SIZE(env->fpr); i++) {
+ if (__get_user(env->fpr[i], &frame->mc_fregs[i])) {
+ return 1;
+ }
+ }
+ if (__get_user(fpscr, &frame->mc_fregs[32]))
+ return 1;
+ env->fpscr = (uint32_t) fpscr;
+ }
+
+ /* Save SPE registers. The kernel only saves the high half. */
+ if (env->insns_flags & PPC_SPE) {
+#if defined(TARGET_PPC64)
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ uint32_t hi;
+
+ if (__get_user(hi, &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ env->gpr[i] = ((uint64_t)hi << 32) | ((uint32_t) env->gpr[i]);
+ }
+#else
+ for (i = 0; i < ARRAY_SIZE(env->gprh); i++) {
+ if (__get_user(env->gprh[i], &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ }
+#endif
+ if (__get_user(env->spe_fscr, &frame->mc_vregs.spe[32]))
+ return 1;
+ }
+
+ return 0;
+}
+
+static void setup_frame(int sig, struct target_sigaction *ka,
+ target_sigset_t *set, CPUState *env)
+{
+ struct target_sigframe *frame;
+ struct target_sigcontext *sc;
+ target_ulong frame_addr, newsp;
+ int err = 0;
+ int signal;
+
+ frame_addr = get_sigframe(ka, env, sizeof(*frame));
+ if (!lock_user_struct(VERIFY_WRITE, frame, frame_addr, 1))
+ goto sigsegv;
+ sc = &frame->sctx;
+
+ signal = current_exec_domain_sig(sig);
+
+ err |= __put_user(h2g(ka->_sa_handler), &sc->handler);
+ err |= __put_user(set->sig[0], &sc->oldmask);
+#if defined(TARGET_PPC64)
+ err |= __put_user(set->sig[0] >> 32, &sc->_unused[3]);
+#else
+ err |= __put_user(set->sig[1], &sc->_unused[3]);
+#endif
+ err |= __put_user(h2g(&frame->mctx), &sc->regs);
+ err |= __put_user(sig, &sc->signal);
+
+ /* Save user regs. */
+ err |= save_user_regs(env, &frame->mctx, TARGET_NR_sigreturn);
+
+ /* The kernel checks for the presence of a VDSO here. We don't
+ emulate a vdso, so use a sigreturn system call. */
+ env->lr = (target_ulong) h2g(frame->mctx.tramp);
+
+ /* Turn off all fp exceptions. */
+ env->fpscr = 0;
+
+ /* Create a stack frame for the caller of the handler. */
+ newsp = frame_addr - SIGNAL_FRAMESIZE;
+ err |= __put_user(env->gpr[1], (target_ulong *)(uintptr_t) newsp);
+
+ if (err)
+ goto sigsegv;
+
+ /* Set up registers for signal handler. */
+ env->gpr[1] = newsp;
+ env->gpr[3] = signal;
+ env->gpr[4] = (target_ulong) h2g(sc);
+ env->nip = (target_ulong) ka->_sa_handler;
+ /* Signal handlers are entered in big-endian mode. */
+ env->msr &= ~MSR_LE;
+
+ unlock_user_struct(frame, frame_addr, 1);
+ return;
+
+sigsegv:
+ unlock_user_struct(frame, frame_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from setup_frame\n");
+ force_sig(SIGSEGV);
+}
+
+static void setup_rt_frame(int sig, struct target_sigaction *ka,
+ target_siginfo_t *info,
+ target_sigset_t *set, CPUState *env)
+{
+ struct target_rt_sigframe *rt_sf;
+ struct target_mcontext *frame;
+ target_ulong rt_sf_addr, newsp = 0;
+ int i, err = 0;
+ int signal;
+
+ rt_sf_addr = get_sigframe(ka, env, sizeof(*rt_sf));
+ if (!lock_user_struct(VERIFY_WRITE, rt_sf, rt_sf_addr, 1))
+ goto sigsegv;
+
+ signal = current_exec_domain_sig(sig);
+
+ err |= copy_siginfo_to_user(&rt_sf->info, info);
+
+ err |= __put_user(0, &rt_sf->uc.uc_flags);
+ err |= __put_user(0, &rt_sf->uc.uc_link);
+ err |= __put_user((target_ulong)target_sigaltstack_used.ss_sp,
+ &rt_sf->uc.uc_stack.ss_sp);
+ err |= __put_user(sas_ss_flags(env->gpr[1]),
+ &rt_sf->uc.uc_stack.ss_flags);
+ err |= __put_user(target_sigaltstack_used.ss_size,
+ &rt_sf->uc.uc_stack.ss_size);
+ err |= __put_user(h2g (&rt_sf->uc.uc_mcontext),
+ &rt_sf->uc.uc_regs);
+ for(i = 0; i < TARGET_NSIG_WORDS; i++) {
+ err |= __put_user(set->sig[i], &rt_sf->uc.uc_sigmask.sig[i]);
+ }
+
+ frame = &rt_sf->uc.uc_mcontext;
+ err |= save_user_regs(env, frame, TARGET_NR_rt_sigreturn);
+
+ /* The kernel checks for the presence of a VDSO here. We don't
+ emulate a vdso, so use a sigreturn system call. */
+ env->lr = (target_ulong) h2g(frame->tramp);
+
+ /* Turn off all fp exceptions. */
+ env->fpscr = 0;
+
+ /* Create a stack frame for the caller of the handler. */
+ newsp = rt_sf_addr - (SIGNAL_FRAMESIZE + 16);
+ err |= __put_user(env->gpr[1], (target_ulong *)(uintptr_t) newsp);
+
+ if (err)
+ goto sigsegv;
+
+ /* Set up registers for signal handler. */
+ env->gpr[1] = newsp;
+ env->gpr[3] = (target_ulong) signal;
+ env->gpr[4] = (target_ulong) h2g(&rt_sf->info);
+ env->gpr[5] = (target_ulong) h2g(&rt_sf->uc);
+ env->gpr[6] = (target_ulong) h2g(rt_sf);
+ env->nip = (target_ulong) ka->_sa_handler;
+ /* Signal handlers are entered in big-endian mode. */
+ env->msr &= ~MSR_LE;
+
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ return;
+
+sigsegv:
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from setup_rt_frame\n");
+ force_sig(SIGSEGV);
+
+}
+
+long do_sigreturn(CPUState *env)
+{
+ struct target_sigcontext *sc = NULL;
+ struct target_mcontext *sr = NULL;
+ target_ulong sr_addr, sc_addr;
+ sigset_t blocked;
+ target_sigset_t set;
+
+ sc_addr = env->gpr[1] + SIGNAL_FRAMESIZE;
+ if (!lock_user_struct(VERIFY_READ, sc, sc_addr, 1))
+ goto sigsegv;
+
+#if defined(TARGET_PPC64)
+ set.sig[0] = sc->oldmask + ((long)(sc->_unused[3]) << 32);
+#else
+ if(__get_user(set.sig[0], &sc->oldmask) ||
+ __get_user(set.sig[1], &sc->_unused[3]))
+ goto sigsegv;
+#endif
+ target_to_host_sigset_internal(&blocked, &set);
+ sigprocmask(SIG_SETMASK, &blocked, NULL);
+
+ if (__get_user(sr_addr, &sc->regs))
+ goto sigsegv;
+ if (!lock_user_struct(VERIFY_READ, sr, sr_addr, 1))
+ goto sigsegv;
+ if (restore_user_regs(env, sr, 1))
+ goto sigsegv;
+
+ unlock_user_struct(sr, sr_addr, 1);
+ unlock_user_struct(sc, sc_addr, 1);
+ return -TARGET_QEMU_ESIGRETURN;
+
+sigsegv:
+ unlock_user_struct(sr, sr_addr, 1);
+ unlock_user_struct(sc, sc_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from do_sigreturn\n");
+ force_sig(SIGSEGV);
+ return 0;
+}
+
+/* See arch/powerpc/kernel/signal_32.c. */
+static int do_setcontext(struct target_ucontext *ucp, CPUState *env, int sig)
+{
+ struct target_mcontext *mcp;
+ target_ulong mcp_addr;
+ sigset_t blocked;
+ target_sigset_t set;
+
+ if (copy_from_user(&set, h2g(ucp) + offsetof(struct target_ucontext, uc_sigmask),
+ sizeof (set)))
+ return 1;
+
+#if defined(TARGET_PPC64)
+ fprintf (stderr, "do_setcontext: not implemented\n");
+ return 0;
+#else
+ if (__get_user(mcp_addr, &ucp->uc_regs))
+ return 1;
+
+ if (!lock_user_struct(VERIFY_READ, mcp, mcp_addr, 1))
+ return 1;
+
+ target_to_host_sigset_internal(&blocked, &set);
+ sigprocmask(SIG_SETMASK, &blocked, NULL);
+ if (restore_user_regs(env, mcp, sig))
+ goto sigsegv;
+
+ unlock_user_struct(mcp, mcp_addr, 1);
+ return 0;
+
+sigsegv:
+ unlock_user_struct(mcp, mcp_addr, 1);
+ return 1;
+#endif
+}
+
+long do_rt_sigreturn(CPUState *env)
+{
+ struct target_rt_sigframe *rt_sf = NULL;
+ target_ulong rt_sf_addr;
+
+ rt_sf_addr = env->gpr[1] + SIGNAL_FRAMESIZE + 16;
+ if (!lock_user_struct(VERIFY_READ, rt_sf, rt_sf_addr, 1))
+ goto sigsegv;
+
+ if (do_setcontext(&rt_sf->uc, env, 1))
+ goto sigsegv;
+
+ do_sigaltstack(rt_sf_addr
+ + offsetof(struct target_rt_sigframe, uc.uc_stack),
+ 0, env->gpr[1]);
+
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ return -TARGET_QEMU_ESIGRETURN;
+
+sigsegv:
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from do_rt_sigreturn\n");
+ force_sig(SIGSEGV);
+ return 0;
+}
+
#else
static void setup_frame(int sig, struct target_sigaction *ka,
--
1.6.0.5
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [Qemu-devel] [PATCH 3/3] support ELF_HWCAP for PPPC
2009-05-12 19:26 [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 1/3] target-ppc: expose cpu capability flags Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 2/3] linux-user: ppc signal handling Nathan Froyd
@ 2009-05-12 19:26 ` Nathan Froyd
2009-05-12 23:13 ` [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 malc
3 siblings, 0 replies; 10+ messages in thread
From: Nathan Froyd @ 2009-05-12 19:26 UTC (permalink / raw)
To: qemu-devel
Signed-off-by: Nathan Froyd <froydnj@codesourcery.com>
---
linux-user/elfload.c | 58 ++++++++++++++++++++++++++++++++++++++++++++++++++
1 files changed, 58 insertions(+), 0 deletions(-)
diff --git a/linux-user/elfload.c b/linux-user/elfload.c
index 6de30f4..508cb37 100644
--- a/linux-user/elfload.c
+++ b/linux-user/elfload.c
@@ -300,6 +300,64 @@ static inline void init_thread(struct target_pt_regs *regs, struct image_info *i
#endif
#define ELF_ARCH EM_PPC
+/* Feature masks for the Aux Vector Hardware Capabilities (AT_HWCAP).
+ See arch/powerpc/include/asm/cputable.h. */
+enum {
+ PPC_FEATURE_32 = 0x80000000,
+ PPC_FEATURE_64 = 0x40000000,
+ PPC_FEATURE_601_INSTR = 0x20000000,
+ PPC_FEATURE_HAS_ALTIVEC = 0x10000000,
+ PPC_FEATURE_HAS_FPU = 0x08000000,
+ PPC_FEATURE_HAS_MMU = 0x04000000,
+ PPC_FEATURE_HAS_4xxMAC = 0x02000000,
+ PPC_FEATURE_UNIFIED_CACHE = 0x01000000,
+ PPC_FEATURE_HAS_SPE = 0x00800000,
+ PPC_FEATURE_HAS_EFP_SINGLE = 0x00400000,
+ PPC_FEATURE_HAS_EFP_DOUBLE = 0x00200000,
+ PPC_FEATURE_NO_TB = 0x00100000,
+ PPC_FEATURE_POWER4 = 0x00080000,
+ PPC_FEATURE_POWER5 = 0x00040000,
+ PPC_FEATURE_POWER5_PLUS = 0x00020000,
+ PPC_FEATURE_CELL = 0x00010000,
+ PPC_FEATURE_BOOKE = 0x00008000,
+ PPC_FEATURE_SMT = 0x00004000,
+ PPC_FEATURE_ICACHE_SNOOP = 0x00002000,
+ PPC_FEATURE_ARCH_2_05 = 0x00001000,
+ PPC_FEATURE_PA6T = 0x00000800,
+ PPC_FEATURE_HAS_DFP = 0x00000400,
+ PPC_FEATURE_POWER6_EXT = 0x00000200,
+ PPC_FEATURE_ARCH_2_06 = 0x00000100,
+ PPC_FEATURE_HAS_VSX = 0x00000080,
+ PPC_FEATURE_PSERIES_PERFMON_COMPAT = 0x00000040,
+
+ PPC_FEATURE_TRUE_LE = 0x00000002,
+ PPC_FEATURE_PPC_LE = 0x00000001,
+};
+
+#define ELF_HWCAP get_elf_hwcap()
+
+static uint32_t get_elf_hwcap(void)
+{
+ CPUState *e = thread_env;
+ uint32_t features = 0;
+
+ /* We don't have to be terribly complete here; the high points are
+ Altivec/FP/SPE support. Anything else is just a bonus. */
+#define GET_FEATURE(flag, feature) \
+ do {if (e->insns_flags & flag) features |= feature; } while(0)
+ GET_FEATURE(PPC_64B, PPC_FEATURE_64);
+ GET_FEATURE(PPC_FLOAT, PPC_FEATURE_HAS_FPU);
+ GET_FEATURE(PPC_ALTIVEC, PPC_FEATURE_HAS_ALTIVEC);
+ GET_FEATURE(PPC_SPE, PPC_FEATURE_HAS_SPE);
+ GET_FEATURE(PPC_SPE_SINGLE, PPC_FEATURE_HAS_EFP_SINGLE);
+ GET_FEATURE(PPC_SPE_DOUBLE, PPC_FEATURE_HAS_EFP_DOUBLE);
+ GET_FEATURE(PPC_BOOKE, PPC_FEATURE_BOOKE);
+ GET_FEATURE(PPC_405_MAC, PPC_FEATURE_HAS_4xxMAC);
+#undef GET_FEATURE
+
+ return features;
+}
+
/*
* We need to put in some extra aux table entries to tell glibc what
* the cache block size is, so it can use the dcbz instruction safely.
--
1.6.0.5
^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2
2009-05-12 19:26 [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 Nathan Froyd
` (2 preceding siblings ...)
2009-05-12 19:26 ` [Qemu-devel] [PATCH 3/3] support ELF_HWCAP for PPPC Nathan Froyd
@ 2009-05-12 23:13 ` malc
2009-05-13 2:15 ` Nathan Froyd
3 siblings, 1 reply; 10+ messages in thread
From: malc @ 2009-05-12 23:13 UTC (permalink / raw)
To: Nathan Froyd; +Cc: qemu-devel
On Tue, 12 May 2009, Nathan Froyd wrote:
> This patch series adds signal handling support for 32-bit PPC linux-user
> emulation. It requires a bit of shuffling of target-ppc code to ensure
> that we know what the capabilities of our current CPU are so we know
> which registers to save.
>
> The only change from the previous version, posted here:
>
> http://lists.gnu.org/archive/html/qemu-devel/2009-04/index.html
>
> is that the PT_* constants have been renamed to TARGET_PT_* constants so
> that everything will build OK on PowerPC Linux hosts.
>
> The signal handling code has been tested by an in-house NPTL
> implementation for PPC linux-user emulation. The NPTL implementation
> will be coming along shortly, but the signal handling bits are useful on
> their own.
This one works as is, however the series introduces 3 instances of
trailing whitespace (git complains) and few new tab characters (my
eyes do). Plus while building it (this time with gcc 3.4.6) i noticed
that the warnings around __put_user are gone, but lines 2746 and 2747
of op_helper.c produce (both originate from d9430add, which is also
yours):
warning: array subscript is above array bounds
Boils down to:
#define VSHIFT(suffix, leftp) \
void helper_vs##suffix (ppc_avr_t *r, ppc_avr_t *a, ppc_avr_t *b) \
{ \
int shift = b->u8[LO_IDX*0x15] & 0x7; \
I'd venture a guess that '*15' instead was meant to be there.
--
mailto:av1474@comtv.ru
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH 2/3] linux-user: ppc signal handling
2009-05-12 19:26 ` [Qemu-devel] [PATCH 2/3] linux-user: ppc signal handling Nathan Froyd
@ 2009-05-13 2:13 ` Nathan Froyd
0 siblings, 0 replies; 10+ messages in thread
From: Nathan Froyd @ 2009-05-13 2:13 UTC (permalink / raw)
To: qemu-devel
Implement setup_{,rt_}frame and do_{,rt_}sigreturn for PPC 32-bit. Use
the same TARGET_QEMU_ESIGRETURN hack as for MIPS to avoid clobbering
register state on a sigreturn.
Signed-off-by: Nathan Froyd <froydnj@codesourcery.com>
---
Remove the trailing whitespace and tabs that malc noticed.
linux-user/main.c | 5 +
linux-user/ppc/syscall.h | 3 +
linux-user/signal.c | 596 ++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 604 insertions(+), 0 deletions(-)
diff --git a/linux-user/main.c b/linux-user/main.c
index dc39b05..9c57c17 100644
--- a/linux-user/main.c
+++ b/linux-user/main.c
@@ -1461,6 +1461,11 @@ void cpu_loop(CPUPPCState *env)
ret = do_syscall(env, env->gpr[0], env->gpr[3], env->gpr[4],
env->gpr[5], env->gpr[6], env->gpr[7],
env->gpr[8]);
+ if (ret == (uint32_t)(-TARGET_QEMU_ESIGRETURN)) {
+ /* Returning from a successful sigreturn syscall.
+ Avoid corrupting register state. */
+ break;
+ }
if (ret > (uint32_t)(-515)) {
env->crf[0] |= 0x1;
ret = -ret;
diff --git a/linux-user/ppc/syscall.h b/linux-user/ppc/syscall.h
index a21d61b..d789911 100644
--- a/linux-user/ppc/syscall.h
+++ b/linux-user/ppc/syscall.h
@@ -51,6 +51,9 @@ struct target_revectored_struct {
abi_ulong __map[8]; /* 256 bits */
};
+/* Nasty hack: define a fake errno value for use by sigreturn. */
+#define TARGET_QEMU_ESIGRETURN 255
+
/*
* flags masks
*/
diff --git a/linux-user/signal.c b/linux-user/signal.c
index 742d52a..9c9c7eb 100644
--- a/linux-user/signal.c
+++ b/linux-user/signal.c
@@ -3182,6 +3182,602 @@ long do_rt_sigreturn(CPUState *env)
return -TARGET_ENOSYS;
}
+#elif defined(TARGET_PPC) && !defined(TARGET_PPC64)
+
+/* FIXME: Many of the structures are defined for both PPC and PPC64, but
+ the signal handling is different enough that we haven't implemented
+ support for PPC64 yet. Hence the restriction above.
+
+ There are various #if'd blocks for code for TARGET_PPC64. These
+ blocks should go away so that we can successfully run 32-bit and
+ 64-bit binaries on a QEMU configured for PPC64. */
+
+/* Size of dummy stack frame allocated when calling signal handler.
+ See arch/powerpc/include/asm/ptrace.h. */
+#if defined(TARGET_PPC64)
+#define SIGNAL_FRAMESIZE 128
+#else
+#define SIGNAL_FRAMESIZE 64
+#endif
+
+/* See arch/powerpc/include/asm/sigcontext.h. */
+struct target_sigcontext {
+ target_ulong _unused[4];
+ int32_t signal;
+#if defined(TARGET_PPC64)
+ int32_t pad0;
+#endif
+ target_ulong handler;
+ target_ulong oldmask;
+ target_ulong regs; /* struct pt_regs __user * */
+ /* TODO: PPC64 includes extra bits here. */
+};
+
+/* Indices for target_mcontext.mc_gregs, below.
+ See arch/powerpc/include/asm/ptrace.h for details. */
+enum {
+ TARGET_PT_R0 = 0,
+ TARGET_PT_R1 = 1,
+ TARGET_PT_R2 = 2,
+ TARGET_PT_R3 = 3,
+ TARGET_PT_R4 = 4,
+ TARGET_PT_R5 = 5,
+ TARGET_PT_R6 = 6,
+ TARGET_PT_R7 = 7,
+ TARGET_PT_R8 = 8,
+ TARGET_PT_R9 = 9,
+ TARGET_PT_R10 = 10,
+ TARGET_PT_R11 = 11,
+ TARGET_PT_R12 = 12,
+ TARGET_PT_R13 = 13,
+ TARGET_PT_R14 = 14,
+ TARGET_PT_R15 = 15,
+ TARGET_PT_R16 = 16,
+ TARGET_PT_R17 = 17,
+ TARGET_PT_R18 = 18,
+ TARGET_PT_R19 = 19,
+ TARGET_PT_R20 = 20,
+ TARGET_PT_R21 = 21,
+ TARGET_PT_R22 = 22,
+ TARGET_PT_R23 = 23,
+ TARGET_PT_R24 = 24,
+ TARGET_PT_R25 = 25,
+ TARGET_PT_R26 = 26,
+ TARGET_PT_R27 = 27,
+ TARGET_PT_R28 = 28,
+ TARGET_PT_R29 = 29,
+ TARGET_PT_R30 = 30,
+ TARGET_PT_R31 = 31,
+ TARGET_PT_NIP = 32,
+ TARGET_PT_MSR = 33,
+ TARGET_PT_ORIG_R3 = 34,
+ TARGET_PT_CTR = 35,
+ TARGET_PT_LNK = 36,
+ TARGET_PT_XER = 37,
+ TARGET_PT_CCR = 38,
+ /* Yes, there are two registers with #39. One is 64-bit only. */
+ TARGET_PT_MQ = 39,
+ TARGET_PT_SOFTE = 39,
+ TARGET_PT_TRAP = 40,
+ TARGET_PT_DAR = 41,
+ TARGET_PT_DSISR = 42,
+ TARGET_PT_RESULT = 43,
+ TARGET_PT_REGS_COUNT = 44
+};
+
+/* See arch/powerpc/include/asm/ucontext.h. Only used for 32-bit PPC;
+ on 64-bit PPC, sigcontext and mcontext are one and the same. */
+struct target_mcontext {
+ target_ulong mc_gregs[48];
+ /* Includes fpscr. */
+ uint64_t mc_fregs[33];
+ target_ulong mc_pad[2];
+ /* We need to handle Altivec and SPE at the same time, which no
+ kernel needs to do. Fortunately, the kernel defines this bit to
+ be Altivec-register-large all the time, rather than trying to
+ twiddle it based on the specific platform. */
+ union {
+ /* SPE vector registers. One extra for SPEFSCR. */
+ uint32_t spe[33];
+ /* Altivec vector registers. The packing of VSCR and VRSAVE
+ varies depending on whether we're PPC64 or not: PPC64 splits
+ them apart; PPC32 stuffs them together. */
+#if defined(TARGET_PPC64)
+#define NVRREG 34
+#else
+#define NVRREG 33
+#endif
+ ppc_avr_t altivec[NVRREG];
+#undef NVRREG
+ } mc_vregs __attribute__((__aligned__(16)));
+};
+
+struct target_ucontext {
+ target_ulong uc_flags;
+ target_ulong uc_link; /* struct ucontext __user * */
+ struct target_sigaltstack uc_stack;
+#if !defined(TARGET_PPC64)
+ int32_t uc_pad[7];
+ target_ulong uc_regs; /* struct mcontext __user *
+ points to uc_mcontext field */
+#endif
+ target_sigset_t uc_sigmask;
+#if defined(TARGET_PPC64)
+ target_sigset_t unused[15]; /* Allow for uc_sigmask growth */
+ struct target_sigcontext uc_mcontext;
+#else
+ int32_t uc_maskext[30];
+ int32_t uc_pad2[3];
+ struct target_mcontext uc_mcontext;
+#endif
+};
+
+/* See arch/powerpc/kernel/signal_32.c. */
+struct target_sigframe {
+ struct target_sigcontext sctx;
+ struct target_mcontext mctx;
+ int32_t abigap[56];
+};
+
+struct target_rt_sigframe {
+ struct target_siginfo info;
+ struct target_ucontext uc;
+ int32_t abigap[56];
+};
+
+/* We use the mc_pad field for the signal return trampoline. */
+#define tramp mc_pad
+
+/* See arch/powerpc/kernel/signal.c. */
+static target_ulong get_sigframe(struct target_sigaction *ka,
+ CPUState *env,
+ int frame_size)
+{
+ target_ulong oldsp, newsp;
+
+ oldsp = env->gpr[1];
+
+ if ((ka->sa_flags & TARGET_SA_ONSTACK) &&
+ (sas_ss_flags(oldsp))) {
+ oldsp = (target_sigaltstack_used.ss_sp
+ + target_sigaltstack_used.ss_size);
+ }
+
+ newsp = (oldsp - frame_size) & ~0xFUL;
+
+ return newsp;
+}
+
+static int save_user_regs(CPUState *env, struct target_mcontext *frame,
+ int sigret)
+{
+ target_ulong msr = env->msr;
+ int i;
+ target_ulong ccr = 0;
+
+ /* In general, the kernel attempts to be intelligent about what it
+ needs to save for Altivec/FP/SPE registers. We don't care that
+ much, so we just go ahead and save everything. */
+
+ /* Save general registers. */
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ if (__put_user(env->gpr[i], &frame->mc_gregs[i])) {
+ return 1;
+ }
+ }
+ if (__put_user(env->nip, &frame->mc_gregs[TARGET_PT_NIP])
+ || __put_user(env->ctr, &frame->mc_gregs[TARGET_PT_CTR])
+ || __put_user(env->lr, &frame->mc_gregs[TARGET_PT_LNK])
+ || __put_user(env->xer, &frame->mc_gregs[TARGET_PT_XER]))
+ return 1;
+
+ for (i = 0; i < ARRAY_SIZE(env->crf); i++) {
+ ccr |= env->crf[i] << (32 - ((i + 1) * 4));
+ }
+ if (__put_user(ccr, &frame->mc_gregs[TARGET_PT_CCR]))
+ return 1;
+
+ /* Save Altivec registers if necessary. */
+ if (env->insns_flags & PPC_ALTIVEC) {
+ for (i = 0; i < ARRAY_SIZE(env->avr); i++) {
+ ppc_avr_t *avr = &env->avr[i];
+ ppc_avr_t *vreg = &frame->mc_vregs.altivec[i];
+
+ if (__put_user(avr->u64[0], &vreg->u64[0]) ||
+ __put_user(avr->u64[1], &vreg->u64[1])) {
+ return 1;
+ }
+ }
+ /* Set MSR_VR in the saved MSR value to indicate that
+ frame->mc_vregs contains valid data. */
+ msr |= MSR_VR;
+ if (__put_user((uint32_t)env->spr[SPR_VRSAVE],
+ &frame->mc_vregs.altivec[32].u32[3]))
+ return 1;
+ }
+
+ /* Save floating point registers. */
+ if (env->insns_flags & PPC_FLOAT) {
+ for (i = 0; i < ARRAY_SIZE(env->fpr); i++) {
+ if (__put_user(env->fpr[i], &frame->mc_fregs[i])) {
+ return 1;
+ }
+ }
+ if (__put_user((uint64_t) env->fpscr, &frame->mc_fregs[32]))
+ return 1;
+ }
+
+ /* Save SPE registers. The kernel only saves the high half. */
+ if (env->insns_flags & PPC_SPE) {
+#if defined(TARGET_PPC64)
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ if (__put_user(env->gpr[i] >> 32, &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ }
+#else
+ for (i = 0; i < ARRAY_SIZE(env->gprh); i++) {
+ if (__put_user(env->gprh[i], &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ }
+#endif
+ /* Set MSR_SPE in the saved MSR value to indicate that
+ frame->mc_vregs contains valid data. */
+ msr |= MSR_SPE;
+ if (__put_user(env->spe_fscr, &frame->mc_vregs.spe[32]))
+ return 1;
+ }
+
+ /* Store MSR. */
+ if (__put_user(msr, &frame->mc_gregs[TARGET_PT_MSR]))
+ return 1;
+
+ /* Set up the sigreturn trampoline: li r0,sigret; sc. */
+ if (sigret) {
+ if (__put_user(0x38000000UL | sigret, &frame->tramp[0]) ||
+ __put_user(0x44000002UL, &frame->tramp[1])) {
+ return 1;
+ }
+ }
+
+ return 0;
+}
+
+static int restore_user_regs(CPUState *env,
+ struct target_mcontext *frame, int sig)
+{
+ target_ulong save_r2 = 0;
+ target_ulong msr;
+ target_ulong ccr;
+
+ int i;
+
+ if (!sig) {
+ save_r2 = env->gpr[2];
+ }
+
+ /* Restore general registers. */
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ if (__get_user(env->gpr[i], &frame->mc_gregs[i])) {
+ return 1;
+ }
+ }
+ if (__get_user(env->nip, &frame->mc_gregs[TARGET_PT_NIP])
+ || __get_user(env->ctr, &frame->mc_gregs[TARGET_PT_CTR])
+ || __get_user(env->lr, &frame->mc_gregs[TARGET_PT_LNK])
+ || __get_user(env->xer, &frame->mc_gregs[TARGET_PT_XER]))
+ return 1;
+ if (__get_user(ccr, &frame->mc_gregs[TARGET_PT_CCR]))
+ return 1;
+
+ for (i = 0; i < ARRAY_SIZE(env->crf); i++) {
+ env->crf[i] = (ccr >> (32 - ((i + 1) * 4))) & 0xf;
+ }
+
+ if (!sig) {
+ env->gpr[2] = save_r2;
+ }
+ /* Restore MSR. */
+ if (__get_user(msr, &frame->mc_gregs[TARGET_PT_MSR]))
+ return 1;
+
+ /* If doing signal return, restore the previous little-endian mode. */
+ if (sig)
+ env->msr = (env->msr & ~MSR_LE) | (msr & MSR_LE);
+
+ /* Restore Altivec registers if necessary. */
+ if (env->insns_flags & PPC_ALTIVEC) {
+ for (i = 0; i < ARRAY_SIZE(env->avr); i++) {
+ ppc_avr_t *avr = &env->avr[i];
+ ppc_avr_t *vreg = &frame->mc_vregs.altivec[i];
+
+ if (__get_user(avr->u64[0], &vreg->u64[0]) ||
+ __get_user(avr->u64[1], &vreg->u64[1])) {
+ return 1;
+ }
+ }
+ /* Set MSR_VEC in the saved MSR value to indicate that
+ frame->mc_vregs contains valid data. */
+ if (__get_user(env->spr[SPR_VRSAVE],
+ (target_ulong *)(&frame->mc_vregs.altivec[32].u32[3])))
+ return 1;
+ }
+
+ /* Restore floating point registers. */
+ if (env->insns_flags & PPC_FLOAT) {
+ uint64_t fpscr;
+ for (i = 0; i < ARRAY_SIZE(env->fpr); i++) {
+ if (__get_user(env->fpr[i], &frame->mc_fregs[i])) {
+ return 1;
+ }
+ }
+ if (__get_user(fpscr, &frame->mc_fregs[32]))
+ return 1;
+ env->fpscr = (uint32_t) fpscr;
+ }
+
+ /* Save SPE registers. The kernel only saves the high half. */
+ if (env->insns_flags & PPC_SPE) {
+#if defined(TARGET_PPC64)
+ for (i = 0; i < ARRAY_SIZE(env->gpr); i++) {
+ uint32_t hi;
+
+ if (__get_user(hi, &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ env->gpr[i] = ((uint64_t)hi << 32) | ((uint32_t) env->gpr[i]);
+ }
+#else
+ for (i = 0; i < ARRAY_SIZE(env->gprh); i++) {
+ if (__get_user(env->gprh[i], &frame->mc_vregs.spe[i])) {
+ return 1;
+ }
+ }
+#endif
+ if (__get_user(env->spe_fscr, &frame->mc_vregs.spe[32]))
+ return 1;
+ }
+
+ return 0;
+}
+
+static void setup_frame(int sig, struct target_sigaction *ka,
+ target_sigset_t *set, CPUState *env)
+{
+ struct target_sigframe *frame;
+ struct target_sigcontext *sc;
+ target_ulong frame_addr, newsp;
+ int err = 0;
+ int signal;
+
+ frame_addr = get_sigframe(ka, env, sizeof(*frame));
+ if (!lock_user_struct(VERIFY_WRITE, frame, frame_addr, 1))
+ goto sigsegv;
+ sc = &frame->sctx;
+
+ signal = current_exec_domain_sig(sig);
+
+ err |= __put_user(h2g(ka->_sa_handler), &sc->handler);
+ err |= __put_user(set->sig[0], &sc->oldmask);
+#if defined(TARGET_PPC64)
+ err |= __put_user(set->sig[0] >> 32, &sc->_unused[3]);
+#else
+ err |= __put_user(set->sig[1], &sc->_unused[3]);
+#endif
+ err |= __put_user(h2g(&frame->mctx), &sc->regs);
+ err |= __put_user(sig, &sc->signal);
+
+ /* Save user regs. */
+ err |= save_user_regs(env, &frame->mctx, TARGET_NR_sigreturn);
+
+ /* The kernel checks for the presence of a VDSO here. We don't
+ emulate a vdso, so use a sigreturn system call. */
+ env->lr = (target_ulong) h2g(frame->mctx.tramp);
+
+ /* Turn off all fp exceptions. */
+ env->fpscr = 0;
+
+ /* Create a stack frame for the caller of the handler. */
+ newsp = frame_addr - SIGNAL_FRAMESIZE;
+ err |= __put_user(env->gpr[1], (target_ulong *)(uintptr_t) newsp);
+
+ if (err)
+ goto sigsegv;
+
+ /* Set up registers for signal handler. */
+ env->gpr[1] = newsp;
+ env->gpr[3] = signal;
+ env->gpr[4] = (target_ulong) h2g(sc);
+ env->nip = (target_ulong) ka->_sa_handler;
+ /* Signal handlers are entered in big-endian mode. */
+ env->msr &= ~MSR_LE;
+
+ unlock_user_struct(frame, frame_addr, 1);
+ return;
+
+sigsegv:
+ unlock_user_struct(frame, frame_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from setup_frame\n");
+ force_sig(SIGSEGV);
+}
+
+static void setup_rt_frame(int sig, struct target_sigaction *ka,
+ target_siginfo_t *info,
+ target_sigset_t *set, CPUState *env)
+{
+ struct target_rt_sigframe *rt_sf;
+ struct target_mcontext *frame;
+ target_ulong rt_sf_addr, newsp = 0;
+ int i, err = 0;
+ int signal;
+
+ rt_sf_addr = get_sigframe(ka, env, sizeof(*rt_sf));
+ if (!lock_user_struct(VERIFY_WRITE, rt_sf, rt_sf_addr, 1))
+ goto sigsegv;
+
+ signal = current_exec_domain_sig(sig);
+
+ err |= copy_siginfo_to_user(&rt_sf->info, info);
+
+ err |= __put_user(0, &rt_sf->uc.uc_flags);
+ err |= __put_user(0, &rt_sf->uc.uc_link);
+ err |= __put_user((target_ulong)target_sigaltstack_used.ss_sp,
+ &rt_sf->uc.uc_stack.ss_sp);
+ err |= __put_user(sas_ss_flags(env->gpr[1]),
+ &rt_sf->uc.uc_stack.ss_flags);
+ err |= __put_user(target_sigaltstack_used.ss_size,
+ &rt_sf->uc.uc_stack.ss_size);
+ err |= __put_user(h2g (&rt_sf->uc.uc_mcontext),
+ &rt_sf->uc.uc_regs);
+ for(i = 0; i < TARGET_NSIG_WORDS; i++) {
+ err |= __put_user(set->sig[i], &rt_sf->uc.uc_sigmask.sig[i]);
+ }
+
+ frame = &rt_sf->uc.uc_mcontext;
+ err |= save_user_regs(env, frame, TARGET_NR_rt_sigreturn);
+
+ /* The kernel checks for the presence of a VDSO here. We don't
+ emulate a vdso, so use a sigreturn system call. */
+ env->lr = (target_ulong) h2g(frame->tramp);
+
+ /* Turn off all fp exceptions. */
+ env->fpscr = 0;
+
+ /* Create a stack frame for the caller of the handler. */
+ newsp = rt_sf_addr - (SIGNAL_FRAMESIZE + 16);
+ err |= __put_user(env->gpr[1], (target_ulong *)(uintptr_t) newsp);
+
+ if (err)
+ goto sigsegv;
+
+ /* Set up registers for signal handler. */
+ env->gpr[1] = newsp;
+ env->gpr[3] = (target_ulong) signal;
+ env->gpr[4] = (target_ulong) h2g(&rt_sf->info);
+ env->gpr[5] = (target_ulong) h2g(&rt_sf->uc);
+ env->gpr[6] = (target_ulong) h2g(rt_sf);
+ env->nip = (target_ulong) ka->_sa_handler;
+ /* Signal handlers are entered in big-endian mode. */
+ env->msr &= ~MSR_LE;
+
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ return;
+
+sigsegv:
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from setup_rt_frame\n");
+ force_sig(SIGSEGV);
+
+}
+
+long do_sigreturn(CPUState *env)
+{
+ struct target_sigcontext *sc = NULL;
+ struct target_mcontext *sr = NULL;
+ target_ulong sr_addr, sc_addr;
+ sigset_t blocked;
+ target_sigset_t set;
+
+ sc_addr = env->gpr[1] + SIGNAL_FRAMESIZE;
+ if (!lock_user_struct(VERIFY_READ, sc, sc_addr, 1))
+ goto sigsegv;
+
+#if defined(TARGET_PPC64)
+ set.sig[0] = sc->oldmask + ((long)(sc->_unused[3]) << 32);
+#else
+ if(__get_user(set.sig[0], &sc->oldmask) ||
+ __get_user(set.sig[1], &sc->_unused[3]))
+ goto sigsegv;
+#endif
+ target_to_host_sigset_internal(&blocked, &set);
+ sigprocmask(SIG_SETMASK, &blocked, NULL);
+
+ if (__get_user(sr_addr, &sc->regs))
+ goto sigsegv;
+ if (!lock_user_struct(VERIFY_READ, sr, sr_addr, 1))
+ goto sigsegv;
+ if (restore_user_regs(env, sr, 1))
+ goto sigsegv;
+
+ unlock_user_struct(sr, sr_addr, 1);
+ unlock_user_struct(sc, sc_addr, 1);
+ return -TARGET_QEMU_ESIGRETURN;
+
+sigsegv:
+ unlock_user_struct(sr, sr_addr, 1);
+ unlock_user_struct(sc, sc_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from do_sigreturn\n");
+ force_sig(SIGSEGV);
+ return 0;
+}
+
+/* See arch/powerpc/kernel/signal_32.c. */
+static int do_setcontext(struct target_ucontext *ucp, CPUState *env, int sig)
+{
+ struct target_mcontext *mcp;
+ target_ulong mcp_addr;
+ sigset_t blocked;
+ target_sigset_t set;
+
+ if (copy_from_user(&set, h2g(ucp) + offsetof(struct target_ucontext, uc_sigmask),
+ sizeof (set)))
+ return 1;
+
+#if defined(TARGET_PPC64)
+ fprintf (stderr, "do_setcontext: not implemented\n");
+ return 0;
+#else
+ if (__get_user(mcp_addr, &ucp->uc_regs))
+ return 1;
+
+ if (!lock_user_struct(VERIFY_READ, mcp, mcp_addr, 1))
+ return 1;
+
+ target_to_host_sigset_internal(&blocked, &set);
+ sigprocmask(SIG_SETMASK, &blocked, NULL);
+ if (restore_user_regs(env, mcp, sig))
+ goto sigsegv;
+
+ unlock_user_struct(mcp, mcp_addr, 1);
+ return 0;
+
+sigsegv:
+ unlock_user_struct(mcp, mcp_addr, 1);
+ return 1;
+#endif
+}
+
+long do_rt_sigreturn(CPUState *env)
+{
+ struct target_rt_sigframe *rt_sf = NULL;
+ target_ulong rt_sf_addr;
+
+ rt_sf_addr = env->gpr[1] + SIGNAL_FRAMESIZE + 16;
+ if (!lock_user_struct(VERIFY_READ, rt_sf, rt_sf_addr, 1))
+ goto sigsegv;
+
+ if (do_setcontext(&rt_sf->uc, env, 1))
+ goto sigsegv;
+
+ do_sigaltstack(rt_sf_addr
+ + offsetof(struct target_rt_sigframe, uc.uc_stack),
+ 0, env->gpr[1]);
+
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ return -TARGET_QEMU_ESIGRETURN;
+
+sigsegv:
+ unlock_user_struct(rt_sf, rt_sf_addr, 1);
+ if (logfile)
+ fprintf (logfile, "segfaulting from do_rt_sigreturn\n");
+ force_sig(SIGSEGV);
+ return 0;
+}
+
#else
static void setup_frame(int sig, struct target_sigaction *ka,
--
1.6.0.5
^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2
2009-05-12 23:13 ` [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 malc
@ 2009-05-13 2:15 ` Nathan Froyd
2009-05-13 11:27 ` malc
0 siblings, 1 reply; 10+ messages in thread
From: Nathan Froyd @ 2009-05-13 2:15 UTC (permalink / raw)
To: qemu-devel
On Wed, May 13, 2009 at 03:13:10AM +0400, malc wrote:
> On Tue, 12 May 2009, Nathan Froyd wrote:
> > This patch series adds signal handling support for 32-bit PPC linux-user
> > emulation. It requires a bit of shuffling of target-ppc code to ensure
> > that we know what the capabilities of our current CPU are so we know
> > which registers to save.
>
> This one works as is, however the series introduces 3 instances of
> trailing whitespace (git complains) and few new tab characters (my
> eyes do).
Sorry about that; fixed in a reply to part 2/3 of the patch series.
> Plus while building it (this time with gcc 3.4.6) i noticed
> that the warnings around __put_user are gone, but lines 2746 and 2747
> of op_helper.c produce (both originate from d9430add, which is also
> yours):
>
> warning: array subscript is above array bounds
>
> Boils down to:
>
> #define VSHIFT(suffix, leftp) \
> void helper_vs##suffix (ppc_avr_t *r, ppc_avr_t *a, ppc_avr_t *b) \
> { \
> int shift = b->u8[LO_IDX*0x15] & 0x7; \
>
> I'd venture a guess that '*15' instead was meant to be there.
Indeed. I will fix this in a separate patch unless you beat me to
it. :) Thanks for the heads-up.
-Nathan
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2
2009-05-13 2:15 ` Nathan Froyd
@ 2009-05-13 11:27 ` malc
2009-05-13 11:37 ` Laurent Desnogues
0 siblings, 1 reply; 10+ messages in thread
From: malc @ 2009-05-13 11:27 UTC (permalink / raw)
To: Nathan Froyd; +Cc: qemu-devel
On Tue, 12 May 2009, Nathan Froyd wrote:
> On Wed, May 13, 2009 at 03:13:10AM +0400, malc wrote:
> > On Tue, 12 May 2009, Nathan Froyd wrote:
> > > This patch series adds signal handling support for 32-bit PPC linux-user
> > > emulation. It requires a bit of shuffling of target-ppc code to ensure
> > > that we know what the capabilities of our current CPU are so we know
> > > which registers to save.
> >
> > This one works as is, however the series introduces 3 instances of
> > trailing whitespace (git complains) and few new tab characters (my
> > eyes do).
>
> Sorry about that; fixed in a reply to part 2/3 of the patch series.
>
> > Plus while building it (this time with gcc 3.4.6) i noticed
> > that the warnings around __put_user are gone, but lines 2746 and 2747
> > of op_helper.c produce (both originate from d9430add, which is also
> > yours):
> >
> > warning: array subscript is above array bounds
> >
> > Boils down to:
> >
> > #define VSHIFT(suffix, leftp) \
> > void helper_vs##suffix (ppc_avr_t *r, ppc_avr_t *a, ppc_avr_t *b) \
> > { \
> > int shift = b->u8[LO_IDX*0x15] & 0x7; \
> >
> > I'd venture a guess that '*15' instead was meant to be there.
>
> Indeed. I will fix this in a separate patch unless you beat me to
> it. :) Thanks for the heads-up.
Done. BTW. here's another:
target-ppc/op_helper.c:1977: warning: comparison is always false due to
limited range of data type
This one is probably harmless, but still.. [The code is due to Aurelien
this time around though]
--
mailto:av1474@comtv.ru
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2
2009-05-13 11:27 ` malc
@ 2009-05-13 11:37 ` Laurent Desnogues
2009-05-13 11:52 ` malc
0 siblings, 1 reply; 10+ messages in thread
From: Laurent Desnogues @ 2009-05-13 11:37 UTC (permalink / raw)
To: malc; +Cc: qemu-devel, Nathan Froyd
On Wed, May 13, 2009 at 1:27 PM, malc <av1474@comtv.ru> wrote:
>
> Done. BTW. here's another:
>
> target-ppc/op_helper.c:1977: warning: comparison is always false due to
> limited range of data type
>
> This one is probably harmless, but still.. [The code is due to Aurelien
> this time around though]
Are you sure? I thought it was part of a patch sent by Nathan.
Anyway that indeed is harmless since it's due to this test (after
macro expansion):
if (0 && x < 0) {
BTW gcc 4.3.2 doesn't complain about it.
Laurent
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2
2009-05-13 11:37 ` Laurent Desnogues
@ 2009-05-13 11:52 ` malc
0 siblings, 0 replies; 10+ messages in thread
From: malc @ 2009-05-13 11:52 UTC (permalink / raw)
To: Laurent Desnogues; +Cc: qemu-devel, Nathan Froyd
On Wed, 13 May 2009, Laurent Desnogues wrote:
> On Wed, May 13, 2009 at 1:27 PM, malc <av1474@comtv.ru> wrote:
> >
> > Done. BTW. here's another:
> >
> > target-ppc/op_helper.c:1977: warning: comparison is always false due to
> > limited range of data type
> >
> > This one is probably harmless, but still.. [The code is due to Aurelien
> > this time around though]
>
> Are you sure? I thought it was part of a patch sent by Nathan.
$ git show `git blame -L 1977 target-ppc/op_helper.c | head -1 | cut -d' ' -f1` | head -10
commit 00d3b8f59bb58e4c7ae52fc73d55d2f27e664443
Author: aurel32 <aurel32@c046a42c-6fe2-441c-8c8c-71466251a162>
Date: Sun Jan 4 22:11:59 2009 +0000
Add saturating arithmetic conversion functions for subsequent
instructions.
Signed-off-by: Nathan Froyd <froydnj@codesourcery.com>
Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@6181
c046a42c-6fe2-441c-8c8c-71466251a162
So yeah you are right it's Nathan's after all.
> Anyway that indeed is harmless since it's due to this test (after
> macro expansion):
>
> if (0 && x < 0) {
>
> BTW gcc 4.3.2 doesn't complain about it.
Looks that way.
--
mailto:av1474@comtv.ru
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2009-05-13 11:52 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-05-12 19:26 [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 1/3] target-ppc: expose cpu capability flags Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 2/3] linux-user: ppc signal handling Nathan Froyd
2009-05-13 2:13 ` Nathan Froyd
2009-05-12 19:26 ` [Qemu-devel] [PATCH 3/3] support ELF_HWCAP for PPPC Nathan Froyd
2009-05-12 23:13 ` [Qemu-devel] [PATCH 0/3] linux-user: add ppc signal handling, v2 malc
2009-05-13 2:15 ` Nathan Froyd
2009-05-13 11:27 ` malc
2009-05-13 11:37 ` Laurent Desnogues
2009-05-13 11:52 ` malc
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.