bsd-user: upstream our signal implementation
PATCH v2 1/40 bsd-user: Complete FreeBSD siginfo
In reply to Warner LoshFill in the missing FreeBSD siginfo fields, and add some comments.
--- bsd-user/freebsd/target_os_siginfo.h | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)
bsd-user/freebsd/target_os_siginfo.h +14 -1
@@ -71,11 +71,24 @@ typedef struct target_siginfo {int32_t _mqd; } _mesgp; -/* SIGPOLL */ +/* SIGPOLL -- Not really genreated in FreeBSD ??? */ struct {int _band;/* POLL_IN, POLL_OUT, POLL_MSG */ } _poll; +struct { +int _mqd; + } _mesgq; + +struct { +/* +* Syscall number for signals delivered as a result of system calls +* denied by Capsicum. +*/ +int _syscall; + } _capsicum; + +/* Spare for future growth */ struct {abi_long __spare1__;int32_t __spare2_[7 ];
Signature · 1 line
2.33.1
PATCH v2 4/40 bsd-user/arm/signal.c: get_mcontext should zero vfp data
In reply to Warner LoshFreeBSD's get_mcontext doesn't return any vfp data. Instead, it zeros out the vfp feilds (and all the spare fields). Impelement this behavior. We're still missing the sysarch(ARM_GET_VFPCONTEXT) syscall, though.
--- bsd-user/arm/signal.c | 8 ++++++++ 1 file changed, 8 insertions(+)
bsd-user/arm/signal.c +8 -0
@@ -109,6 +109,14 @@ abi_long get_mcontext(CPUARMState *env, target_mcontext_t *mcp, int flags) gr[TARGET_REG_LR ] =tswap32 (env->regs[14 ]); gr[TARGET_REG_PC ] =tswap32 (env->regs[15 ]); +/* +* FreeBSD's set_mcontext doesn't save VFP info, but blanks it out instead. +* Instead, sysarch(ARM_GET_VFPSTATE) is used instead. +*/ + mcp->mc_vfp_size =0 ; + mcp->mc_vfp_ptr =0 ; +memset (&mcp->mc_spare,0 ,sizeof (mcp->mc_spare)); +return 0 ; }
Signature · 1 line
2.33.1
PATCH v2 7/40 bsd-user/arm/target_arch_cpu.h: Move EXCP_ATOMIC to match linux-user
In reply to Warner LoshMove the EXCP_ATOMIC case to match linux-user/arm/cpu_loop.c:cpu_loop ordering.
--- bsd-user/arm/target_arch_cpu.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
bsd-user/arm/target_arch_cpu.h +3 -3
@@ -180,12 +180,12 @@ static inline void target_cpu_loop(CPUARMState *env)queue_signal (env, info.si_signo, &info); }break ; -case EXCP_ATOMIC : -cpu_exec_step_atomic (cs); -break ;case EXCP_YIELD :/* nothing to do here for user-mode, just resume guest code */ break ; +case EXCP_ATOMIC : +cpu_exec_step_atomic (cs); +break ;default :fprintf (stderr,"qemu: unhandled CPU exception 0x%x - aborting\n" , trapnr);
Signature · 1 line
2.33.1
v2 cover bsd-user: upstream our signal implementation
Upstream the bsd-user fork signal implementation, for the most part. This series of commits represents nearly all of the infrastructure that surround signals, except the actual system call glue (that was also reworked in the fork and needs its own series). In addition, this adds the sigsegv and sigbus code to arm. Even in the fork, we don't have good x86 signal implementation, so there's little to upstream for that at the moment.
bsd-user's signal implementation is similar to linux-user's. However, all signals are always queued for batch processing (except synchronous ones that kill the process). The full context can be found in the fork's 'blitz branch' at https://github.com/qemu-bsd-user/qemu-bsd-user/tree/blitz which shows how these are used to implement various system calls. Since this was built from linux-user's stack stuff, evolved for BSD with the passage of a few years, it no-doubt missed some bug fixes from linux-user (though nothing obvious stood out in the quick comparison I made).
I lumped thinks slightly larger than past patch sets, but none of the patches should exceed about 100 lines of diffs (there is one that clocks in at 166 though, but I had trouble splitting it smaller). With over 30k lines of diffs between the two repos, I need to find more efficient ways of getting things reviewed and each extra chunk takes time to curate so I'm searching for a good happy medium.
New in version 2:
Lots of changes based on feedback garnered in v1. The patch numbers are only loosely related as some were dropped or merged and others were added. I've not rebased this yet to a newer revision due to the volume of changes.
o Removed an unnecessary memset for the signal table, it's already all 0's.
o Fixed a number of comments as suggested in the first round of review.
o yoda speak removed have I.
o added a path to the MAINTAINERS entry for bsd related VMs
o removed MIPS code from trapframe construction
o Add si_code call to queue_signal
o Remove final remnants of signal queueing on BSD
o Add documentation of fields of the task structure
o Remove a bunch of comments that are no longer correct or turned out to be false...
o Get FSR directly from the siginfo rather than doing crazy things to find it.
o Move target_sigaltstack_used to ts->sigaltstack_used and adjust all references to be per-stack.
o rename force_sig to dump_core_and_abort
o Create a new setup_sigframe_arch to save/restore context. Most architectures this is just get_mcontext, but arm needs special setup.
o Move to generating SIGILL when we can't write the signal trap frame.
o Add comments about a few extreme edge cases for SIGILL instructions
o rewrite arm's host_signal_write to use the trap frame it needs to get the data rather than the prior bogus assumptions.
o Add more comments to the signal related Task elements, cribbed from linux-user
o Move to storing QEMU supecific signal si_code, stored in the top 8 bits of si_code. BSD uses bit 16 to indicate the code is 'generic' and not tied to a specific signal type, so we have to preserve that bit. Only bit 16 is used, but steer clear by using the top 8 bits.
o Don't gratuitously move fatal_signal
o add has_trapno function for those signals that have si_trapno.
o Move to using synchronous signals or signals artificially crated by qemu. Since only one of these can ever happen at the time, we can simplify the code in a manner similar to linux-user.
o Fix a number of "see this routine" pointers to match modern FreeBSD and/or remove obsolete architectures from the generic pointers.
o Fixed the division of labor between host_to_target_siginfo_noswap and tswap_siginfo. We now record a 'si_type' field that can be used to determine which parts of the siginfo are valid. _noswap will record its guess and if that guess is present, tswap_siginfo will use it.
o Fix arm get_mcontext to match what FreeBSD does by removing saving of vfp there and zeroing the vfp pointers.
o Exclude SIGPROF when CONFIG_GPROF active.
o A few formatting issues that I'd ignored for version 1.
o consistently used env as the name for CPUArchState (noted in one patch, but present in more and in a couple places in upstream so I fixed those too).
Not adressed:
o Signal mask issues around sigaction and sigreturn. These will be addressed while people comment on the rest of these changes. I've noted the exact details in a couple of commit messages.
Warner Losh (40):
bsd-user: Complete FreeBSD siginfo
bsd-user: Create setup_sigframe_arch to setup sigframe context bsd-user/arm/signal.c: Implement setup_sigframe_arch for arm bsd-user/arm/signal.c: get_mcontext should zero vfp data
bsd-user: Remove vestiges of signal queueing code
bsd-user: Bring in docs from linux-user for signal_pending bsd-user/arm/target_arch_cpu.h: Move EXCP_ATOMIC to match linux-user bsd-user/signal.c: implement force_sig_fault bsd-user/signal-common.h: Move signal functions prototypes to here bsd-user/signal.c: Implement cpu_loop_exit_sigsegv bsd-user/signal.c: implement cpu_loop_exit_sigbus bsd-user/arm/arget_arch_cpu.h: Move EXCP_DEBUG and EXCP_BKPT together bsd-user/arm/target_arch_cpu.h: Correct code pointer bsd-user/arm/target_arch_cpu.h: Use force_sig_fault for EXCP_UDEF bsd-user/arm/target_arch_cpu.h: Implement data faults bsd-user/signal.c: implement abstract target / host signal translation bsd-user/signal.c: Implement signal_init() bsd-user/signal.c: Add si_type argument to queue_signal bsd-user/host/arm/host-signal.h: Implement host_signal_* bsd-user/host/i386/host-signal.h: Implement host_signal_* bsd-user/host/x86_64/host-signal.h: Implement host_signal_*
bsd-user: Add host signals to the build
bsd-user: Add trace events for bsd-user
bsd-user/signal.c: host_to_target_siginfo_noswap bsd-user/signal.c: Implement rewind_if_in_safe_syscall bsd-user/signal.c: Implement host_signal_handler bsd-user/strace.c: print_taken_signal
bsd-user/signal.c: Implement dump_core_and_abort bsd-user/signal.c: Fill in queue_signal
bsd-user/signal.c: sigset manipulation routines. bsd-user/signal.c: setup_frame
bsd-user/signal.c: handle_pending_signal bsd-user/signal.c: tswap_siginfo
bsd-user/signal.c: process_pending_signals bsd-user/signal.c: implement do_sigreturn bsd-user/signal.c: implement do_sigaction bsd-user/signal.c: do_sigaltstack
MAINTAINERS: Add tests/vm/*bsd to the list to get reviews on
bsd-user: Rename arg name for target_cpu_reset to env bsd-user/freebsd/target_os_ucontext.h: Prefer env as arg name for
CPUArchState args
MAINTAINERS | 1 + bsd-user/arm/signal.c | 58 +- bsd-user/arm/target_arch_cpu.h | 101 +-- bsd-user/freebsd/target_os_siginfo.h | 15 +- bsd-user/freebsd/target_os_signal.h | 3 + bsd-user/freebsd/target_os_ucontext.h | 6 +- bsd-user/host/arm/host-signal.h | 35 + bsd-user/host/i386/host-signal.h | 37 + bsd-user/host/x86_64/host-signal.h | 37 + bsd-user/i386/signal.c | 13 + bsd-user/i386/target_arch_cpu.h | 4 +- bsd-user/main.c | 14 +- bsd-user/meson.build | 1 + bsd-user/qemu.h | 62 +- bsd-user/signal-common.h | 70 ++ bsd-user/signal.c | 1001 ++++++++++++++++++++++++- bsd-user/strace.c | 97 +++ bsd-user/syscall_defs.h | 1 + bsd-user/trace-events | 11 + bsd-user/trace.h | 1 + bsd-user/x86_64/signal.c | 13 + bsd-user/x86_64/target_arch_cpu.h | 4 +- meson.build | 6 +- 23 files changed, 1477 insertions(+), 114 deletions(-) create mode 100644 bsd-user/host/arm/host-signal.h create mode 100644 bsd-user/host/i386/host-signal.h create mode 100644 bsd-user/host/x86_64/host-signal.h create mode 100644 bsd-user/signal-common.h create mode 100644 bsd-user/trace-events create mode 100644 bsd-user/trace.h
Signature · 1 line
2.33.1
PATCH v2 2/40 bsd-user: Create setup_sigframe_arch to setup sigframe context
Define setup_sigframe_arch whose job it is to setup the mcontext for the sigframe. Implement for x86 to just call mcontext.
--- bsd-user/freebsd/target_os_signal.h | 3 +++ bsd-user/i386/signal.c | 13 +++++++++++++ bsd-user/x86_64/signal.c | 13 +++++++++++++ 3 files changed, 29 insertions(+)
bsd-user/freebsd/target_os_signal.h +3 -0
@@ -4,6 +4,9 @@#include "target_os_siginfo.h" #include "target_arch_signal.h" +abi_long setup_sigframe_arch (CPUArchState *regs,abi_ulong frame_addr, +struct target_sigframe *frame,int flags); +/* Compare to sys/signal.h */ #define TARGET_SIGHUP 1/* hangup */ #define TARGET_SIGINT 2/* interrupt */
bsd-user/i386/signal.c +13 -0
@@ -32,6 +32,19 @@ abi_long set_sigtramp_args(CPUX86State *env, int sig,return 0 ; } +/* +* Compare to i386/i386/exec_machdep.c sendsig() +* Assumes that the memory is locked if frame points to user memory. +*/ +abi_long setup_sigframe_arch (CPUX86State *env,abi_ulong frame_addr, +struct target_sigframe *frame,int flags) +{ +target_mcontext_t *mcp = &frame->sf_uc.uc_mcontext; + +get_mcontext (env, mcp, flags); +return 0 ; +} +/* Compare to i386/i386/machdep.c get_mcontext() */ abi_long get_mcontext (CPUX86State *regs,target_mcontext_t *mcp,int flags) {
bsd-user/x86_64/signal.c +13 -0
@@ -30,6 +30,19 @@ abi_long set_sigtramp_args(CPUX86State *regs,return 0 ; } +/* +* Compare to amd64/amd64/exec_machdep.c sendsig() +* Assumes that the memory is locked if frame points to user memory. +*/ +abi_long setup_sigframe_arch (CPUX86State *env,abi_ulong frame_addr, +struct target_sigframe *frame,int flags) +{ +target_mcontext_t *mcp = &frame->sf_uc.uc_mcontext; + +get_mcontext (env, mcp, flags); +return 0 ; +} +/* Compare to amd64/amd64/machdep.c get_mcontext() */ abi_long get_mcontext (CPUX86State *regs,target_mcontext_t *mcp,int flags)
Signature · 1 line
2.33.1
PATCH v2 6/40 bsd-user: Bring in docs from linux-user for signal_pending
In reply to Warner LoshThis is currently unused, so no code adjustments are needed.
--- bsd-user/qemu.h | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
bsd-user/qemu.h +8 -1
@@ -86,7 +86,14 @@ typedef struct TaskState {struct image_info *info;struct emulated_sigtable sigtab[TARGET_NSIG ]; -int signal_pending;/* non zero if a signal may be pending */ +/* +* Nonzero if process_pending_signals() needs to do something (either +* handle a pending signal or unblock signals). +* This flag is written from a signal handler so should be accessed via +* the qatomic_read() and qatomic_set() functions. (It is not accessed +* from multiple threads.) +*/ +int signal_pending;uint8_t stack[]; } __attribute__((aligned(16))) TaskState;
Signature · 1 line
2.33.1
PATCH v2 11/40 bsd-user/signal.c: implement cpu_loop_exit_sigbus
In reply to Warner LoshFirst attempt at implementing cpu_loop_exit_sigbus, mostly copied from linux-user version of this function.
--- bsd-user/signal.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-)
bsd-user/signal.c +9 -3
@@ -80,7 +80,13 @@ void cpu_loop_exit_sigsegv(CPUState *cpu, target_ulong addr,void cpu_loop_exit_sigbus (CPUState *cpu,target_ulong addr,MMUAccessType access_type,uintptr_t ra) { -qemu_log_mask (LOG_UNIMP ,"No signal support for SIGBUS\n" ); -/* unreachable */ -abort (); +const struct TCGCPUOps *tcg_ops =CPU_GET_CLASS (cpu)->tcg_ops; + +if (tcg_ops->record_sigbus) { + tcg_ops->record_sigbus (cpu, addr, access_type, ra); + } + +force_sig_fault (TARGET_SIGBUS ,TARGET_BUS_ADRALN , addr); + cpu->exception_index =EXCP_INTERRUPT ; +cpu_loop_exit_restore (cpu, ra); }
Signature · 1 line
2.33.1
PATCH v2 3/40 bsd-user/arm/signal.c: Implement setup_sigframe_arch for arm
In reply to Warner LoshFix the broken context setting for arm. FreeBSD's get_mcontext does not fill in the vfp info. It's filled in in sigframe(). This corresponds to the new setup_sigframe_arch which fills in mcontext, then adjusts it to point to the vfp context in the sigframe and fills in that context as well. Add pointer to where this code is done.
--- bsd-user/arm/signal.c | 50 +++++++++++++++++++---------- bsd-user/freebsd/target_os_signal.h | 2 +- 2 files changed, 34 insertions(+), 18 deletions(-)
bsd-user/arm/signal.c +33 -17
@@ -59,19 +59,31 @@ abi_long set_sigtramp_args(CPUARMState *env, int sig,return 0 ; } +static abi_long get_vfpcontext (CPUARMState *env,abi_ulong frame_addr, +struct target_sigframe *frame) +{ +/* see sendsig and get_vfpcontext in sys/arm/arm/exec_machdep.c */ +target_mcontext_vfp_t *vfp = &frame->sf_vfp; +target_mcontext_t *mcp = &frame->sf_uc.uc_mcontext; + +/* Assumes that mcp and vfp are locked */ +for (int i =0 ; i <32 ; i++) { + vfp->mcv_reg[i] =tswap64 (*aa32_vfp_dreg (env, i)); + } + vfp->mcv_fpscr =tswap32 (vfp_get_fpscr (env)); + mcp->mc_vfp_size =tswap32 (sizeof (*vfp)); + mcp->mc_vfp_ptr =tswap32 (frame_addr + ((uintptr_t )vfp - (uintptr_t )frame)); +return 0 ; +} +/* -* Compare to arm/arm/machdep.c get_mcontext() +* Compare to arm/arm/exec_machdep.c get_mcontext() * Assumes that the memory is locked if mcp points to user memory. */ abi_long get_mcontext (CPUARMState *env,target_mcontext_t *mcp,int flags) { -int err =0 ;uint32_t *gr = mcp->__gregs; -if (mcp->mc_vfp_size !=0 && mcp->mc_vfp_size !=sizeof (target_mcontext_vfp_t)) { -return -TARGET_EINVAL ; - } - gr[TARGET_REG_CPSR ] =tswap32 (cpsr_read (env));if (flags &TARGET_MC_GET_CLEAR_RET ) { gr[TARGET_REG_R0 ] =0 ;
@@ -97,17 +109,21 @@ abi_long get_mcontext(CPUARMState *env, target_mcontext_t *mcp, int flags) gr[TARGET_REG_LR ] =tswap32 (env->regs[14 ]); gr[TARGET_REG_PC ] =tswap32 (env->regs[15 ]); -if (mcp->mc_vfp_size !=0 && mcp->mc_vfp_ptr !=0 ) { -/* see get_vfpcontext in sys/arm/arm/exec_machdep.c */ -target_mcontext_vfp_t *vfp; - vfp =lock_user (VERIFY_WRITE , mcp->mc_vfp_ptr,sizeof (*vfp),0 ); -for (int i =0 ; i <32 ; i++) { - vfp->mcv_reg[i] =tswap64 (*aa32_vfp_dreg (env, i)); - } - vfp->mcv_fpscr =tswap32 (vfp_get_fpscr (env)); -unlock_user (vfp, mcp->mc_vfp_ptr,sizeof (*vfp)); - } -return err; +return 0 ; +} + +/* +* Compare to arm/arm/exec_machdep.c sendsig() +* Assumes that the memory is locked if frame points to user memory. +*/ +abi_long setup_sigframe_arch (CPUARMState *env,abi_ulong frame_addr, +struct target_sigframe *frame,int flags) +{ +target_mcontext_t *mcp = &frame->sf_uc.uc_mcontext; + +get_mcontext (env, mcp, flags); +get_vfpcontext (env, frame_addr, frame); +return 0 ; }/* Compare to arm/arm/exec_machdep.c set_mcontext() */
bsd-user/freebsd/target_os_signal.h +1 -1
@@ -4,7 +4,7 @@#include "target_os_siginfo.h" #include "target_arch_signal.h" -abi_long setup_sigframe_arch (CPUArchState *regs,abi_ulong frame_addr, +abi_long setup_sigframe_arch (CPUArchState *env,abi_ulong frame_addr,struct target_sigframe *frame,int flags);/* Compare to sys/signal.h */
Signature · 1 line
2.33.1
PATCH v2 5/40 bsd-user: Remove vestiges of signal queueing code
In reply to Warner Loshbsd-user was copied from linux-user at a time when it queued signals. Remove those vestiges of thse code. Retain the init function, even though it's now empty since other stuff will likely be added there. Make it static since it's not called from outside of main.c
--- bsd-user/main.c | 9 +-------- bsd-user/qemu.h | 13 +------------ 2 files changed, 2 insertions(+), 20 deletions(-)
bsd-user/main.c +1 -8
@@ -215,15 +215,8 @@ void qemu_cpu_kick(CPUState *cpu) }/* Assumes contents are already zeroed. */ -void init_task_state (TaskState *ts) +static void init_task_state (TaskState *ts) { -int i; - - ts->first_free = ts->sigqueue_table; -for (i =0 ; i <MAX_SIGQUEUE_SIZE -1 ; i++) { - ts->sigqueue_table[i].next = &ts->sigqueue_table[i +1 ]; - } - ts->sigqueue_table[i].next =NULL ; }void gemu_log (const char *fmt, ...)
bsd-user/qemu.h +1 -12
@@ -70,17 +70,9 @@ struct image_info {uint32_t elf_flags; }; -#define MAX_SIGQUEUE_SIZE 1024 - -struct qemu_sigqueue { -struct qemu_sigqueue *next; -target_siginfo_t info; -}; -struct emulated_sigtable {int pending;/* true if signal is pending */ -struct qemu_sigqueue *first; -struct qemu_sigqueue info;/* Put first signal info here */ +target_siginfo_t info; };/*
@@ -94,14 +86,11 @@ typedef struct TaskState {struct image_info *info;struct emulated_sigtable sigtab[TARGET_NSIG ]; -struct qemu_sigqueue sigqueue_table[MAX_SIGQUEUE_SIZE ];/* siginfo queue */ -struct qemu_sigqueue *first_free;/* first free siginfo queue entry */ int signal_pending;/* non zero if a signal may be pending */ uint8_t stack[]; } __attribute__((aligned(16))) TaskState; -void init_task_state (TaskState *ts);void stop_all_tasks (void );extern const char *qemu_uname_release;
Signature · 1 line
2.33.1
PATCH v2 9/40 bsd-user/signal-common.h: Move signal functions prototypes to here
In reply to Warner Losh--- bsd-user/qemu.h | 8 -------- bsd-user/signal-common.h | 6 ++++++ 2 files changed, 6 insertions(+), 8 deletions(-)
bsd-user/qemu.h +0 -8
@@ -199,14 +199,6 @@ print_openbsd_syscall(int num,void print_openbsd_syscall_ret (int num,abi_long ret);extern int do_strace; -/* signal.c */ -void process_pending_signals (CPUArchState *cpu_env); -void signal_init (void ); -long do_sigreturn (CPUArchState *env); -long do_rt_sigreturn (CPUArchState *env); -void queue_signal (CPUArchState *env,int sig,target_siginfo_t *info); -abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp); -/* mmap.c */ int target_mprotect (abi_ulong start,abi_ulong len,int prot); abi_long target_mmap(abi_ulong start, abi_ulong len, int prot,
bsd-user/signal-common.h +6 -0
@@ -9,6 +9,12 @@#ifndef SIGNAL_COMMON_H #define SIGNAL_COMMON_H +long do_rt_sigreturn (CPUArchState *env); +abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp); +long do_sigreturn (CPUArchState *env);void force_sig_fault (int sig,int code,abi_ulong addr); +void process_pending_signals (CPUArchState *env); +void queue_signal (CPUArchState *env,int sig,target_siginfo_t *info); +void signal_init (void );#endif
Signature · 1 line
2.33.1
PATCH v2 12/40 bsd-user/arm/arget_arch_cpu.h: Move EXCP_DEBUG and EXCP_BKPT together
In reply to Warner LoshImplement EXCP_DEBUG and EXCP_BKPT the same, as is done in linux-user. The prior adjustment of register 15 isn't needed, so remove that. Remove a redunant comment (that code in FreeBSD never handled break points). It's unclear why BKPT was an alias for system calls, but FreeBSD doesn't do that today.
--- bsd-user/arm/target_arch_cpu.h | 23 +++-------------------- 1 file changed, 3 insertions(+), 20 deletions(-)
bsd-user/arm/target_arch_cpu.h +3 -20
@@ -21,6 +21,7 @@#define _TARGET_ARCH_CPU_H_#include "target_arch.h" +#include "signal-common.h" #define TARGET_DEFAULT_CPU_MODEL "any"
@@ -64,19 +65,7 @@ static inline void target_cpu_loop(CPUARMState *env) }break ;case EXCP_SWI : -case EXCP_BKPT : { -/* -* system call -* See arm/arm/trap.c cpu_fetch_syscall_args() -*/ -if (trapnr ==EXCP_BKPT ) { -if (env->thumb) { - env->regs[15 ] +=2 ; - }else { - env->regs[15 ] +=4 ; - } - } n = env->regs[7 ];if (bsd_type == target_freebsd) {int ret;
@@ -171,14 +160,8 @@ static inline void target_cpu_loop(CPUARMState *env)queue_signal (env, info.si_signo, &info);break ;case EXCP_DEBUG : - { - - info.si_signo =TARGET_SIGTRAP ; - info.si_errno =0 ; - info.si_code =TARGET_TRAP_BRKPT ; - info.si_addr = env->exception.vaddress; -queue_signal (env, info.si_signo, &info); - } +case EXCP_BKPT : +force_sig_fault (TARGET_SIGTRAP ,TARGET_TRAP_BRKPT , env->regs[15 ]);break ;case EXCP_YIELD :/* nothing to do here for user-mode, just resume guest code */
Signature · 1 line
2.33.1
PATCH v2 10/40 bsd-user/signal.c: Implement cpu_loop_exit_sigsegv
In reply to Warner LoshFirst attempt at implementing cpu_loop_exit_sigsegv, mostly copied from linux-user version of this function.
--- bsd-user/signal.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-)
bsd-user/signal.c +12 -3
@@ -20,6 +20,7 @@#include "qemu/osdep.h" #include "qemu.h" #include "signal-common.h" +#include "hw/core/tcg-cpu-ops.h" /* * Stubbed out routines until we merge signal support from bsd-user
@@ -63,9 +64,17 @@ void process_pending_signals(CPUArchState *cpu_env)void cpu_loop_exit_sigsegv (CPUState *cpu,target_ulong addr,MMUAccessType access_type,bool maperr,uintptr_t ra) { -qemu_log_mask (LOG_UNIMP ,"No signal support for SIGSEGV\n" ); -/* unreachable */ -abort (); +const struct TCGCPUOps *tcg_ops =CPU_GET_CLASS (cpu)->tcg_ops; + +if (tcg_ops->record_sigsegv) { + tcg_ops->record_sigsegv (cpu, addr, access_type, maperr, ra); + } + +force_sig_fault (TARGET_SIGSEGV , + maperr ?TARGET_SEGV_MAPERR :TARGET_SEGV_ACCERR , + addr); + cpu->exception_index =EXCP_INTERRUPT ; +cpu_loop_exit_restore (cpu, ra); } void cpu_loop_exit_sigbus(CPUState *cpu, target_ulong addr,
Signature · 1 line
2.33.1
PATCH v2 8/40 bsd-user/signal.c: implement force_sig_fault
In reply to Warner LoshStart to implement the force_sig_fault code. This currently just calls queue_signal(). The bsd-user fork version of that will handle this the synchronous nature of this call. Add signal-common.h to hold signal helper functions like force_sig_fault.
--- bsd-user/signal-common.h | 14 ++++++++++++++ bsd-user/signal.c | 18 ++++++++++++++++++ 2 files changed, 32 insertions(+) create mode 100644 bsd-user/signal-common.h
bsd-user/signal-common.h new file +14 -0
@@ -0,0 +1,14 @@ +/* +* Emulation of BSD signals +* +* Copyright (c) 2013 Stacey Son +* +* SPDX-License-Identifier: GPL-2.0-or-later +*/ + +#ifndef SIGNAL_COMMON_H +#define SIGNAL_COMMON_H + +void force_sig_fault (int sig,int code,abi_ulong addr); + +#endif
bsd-user/signal.c +18 -0
@@ -19,6 +19,7 @@#include "qemu/osdep.h" #include "qemu.h" +#include "signal-common.h" /* * Stubbed out routines until we merge signal support from bsd-user
@@ -34,6 +35,23 @@ void queue_signal(CPUArchState *env, int sig, target_siginfo_t *info)qemu_log_mask (LOG_UNIMP ,"No signal queueing, dropping signal %d\n" , sig); } +/* +* Force a synchronously taken QEMU_SI_FAULT signal. For QEMU the +* 'force' part is handled in process_pending_signals(). +*/ +void force_sig_fault (int sig,int code,abi_ulong addr) +{ +CPUState *cpu = thread_cpu; +CPUArchState *env = cpu->env_ptr; +target_siginfo_t info = {}; + + info.si_signo = sig; + info.si_errno =0 ; + info.si_code = code; + info.si_addr = addr; +queue_signal (env, sig, &info); +} +void signal_init (void ) { }
Signature · 1 line
2.33.1
PATCH v2 20/40 bsd-user/host/i386/host-signal.h: Implement host_signal_*
In reply to Warner LoshImplement host_signal_pc, host_signal_set_pc and host_signal_write for i386.
--- bsd-user/host/i386/host-signal.h | 37 ++++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) create mode 100644 bsd-user/host/i386/host-signal.h
bsd-user/host/i386/host-signal.h new file +37 -0
@@ -0,0 +1,37 @@ +/* +* host-signal.h: signal info dependent on the host architecture +* +* Copyright (c) 2021 Warner Losh +* +* SPDX-License-Identifier: GPL-2.0-or-later +*/ + +#ifndef I386_HOST_SIGNAL_H +#define I386_HOST_SIGNAL_H + +#include <sys/ucontext.h> +#include <machine/trap.h> +#include <vm/pmap.h> +#include <machine/pmap.h> + +static inline uintptr_t host_signal_pc (ucontext_t *uc) +{ +return uc->uc_mcontext.mc_eip; +} + +static inline void host_signal_set_pc (ucontext_t *uc,uintptr_t pc) +{ + uc->uc_mcontext.mc_eip = pc; +} + +static inline bool host_signal_write (siginfo_t *info,ucontext_t *uc) +{ +/* +* Look in sys/i386/i386/trap.c. NOTE: mc_err == tr_err due to type punning +* between a trapframe and mcontext on FreeBSD/i386. +*/ +return uc->uc_mcontext.mc_trapno ==T_PAGEFLT && + uc->uc_mcontext.mc_err &PGEX_W ; +} + +#endif
Signature · 1 line
2.33.1
PATCH v2 14/40 bsd-user/arm/target_arch_cpu.h: Use force_sig_fault for EXCP_UDEF
In reply to Warner LoshUse force_sig_fault to implement unknown opcode. This just uninlines that function, so simplify things by using it. Fold in EXCP_NOCP and EXCP_INVSTATE, as is done in linux-user. Make a note about slight differences with FreeBSD in case any of them turn out to be important later.
--- bsd-user/arm/target_arch_cpu.h | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-)
bsd-user/arm/target_arch_cpu.h +13 -12
@@ -51,18 +51,19 @@ static inline void target_cpu_loop(CPUARMState *env)process_queued_cpu_work (cs);switch (trapnr) {case EXCP_UDEF : - { -/* See arm/arm/undefined.c undefinedinstruction(); */ - info.si_addr = env->regs[15 ]; - -/* illegal instruction */ - info.si_signo =TARGET_SIGILL ; - info.si_errno =0 ; - info.si_code =TARGET_ILL_ILLOPC ; -queue_signal (env, info.si_signo, &info); - -/* TODO: What about instruction emulation? */ - } +case EXCP_NOCP : +case EXCP_INVSTATE : +/* +* See arm/arm/undefined.c undefinedinstruction(); +* +* A number of details aren't emulated (they likely don't matter): +* o Misaligned PC generates ILL_ILLADR +* o Thumb-2 instructions generate ILLADR +* o Both modes implement coprocessor instructions, which we don't +* do here. FreeBSD just implements them for the VFP coprocessor +* and special kernel breakpoints, trace points, dtrace, etc. +*/ +force_sig_fault (TARGET_SIGILL ,TARGET_ILL_ILLOPC , env->regs[15 ]);break ;case EXCP_SWI : {
Signature · 1 line
2.33.1
PATCH v2 13/40 bsd-user/arm/target_arch_cpu.h: Correct code pointer
In reply to Warner LoshThe code has moved in FreeBSD since the emulator was started, update the comment to reflect that change.
--- bsd-user/arm/target_arch_cpu.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
bsd-user/arm/target_arch_cpu.h +1 -1
@@ -73,7 +73,7 @@ static inline void target_cpu_loop(CPUARMState *env)int32_t syscall_nr = n;int32_t arg1, arg2, arg3, arg4, arg5, arg6, arg7, arg8; -/* See arm/arm/trap.c cpu_fetch_syscall_args() */ +/* See arm/arm/syscall.c cpu_fetch_syscall_args() */ if (syscall_nr == TARGET_FREEBSD_NR_syscall) { syscall_nr = env->regs[0 ]; arg1 = env->regs[1 ];
Signature · 1 line
2.33.1
PATCH v2 31/40 bsd-user/signal.c: setup_frame
In reply to Warner Loshsetup_frame sets up a signalled stack frame. Associated routines to extract the pointer to the stack frame and to support alternate stacks.
--- bsd-user/main.c | 5 +++ bsd-user/qemu.h | 3 +- bsd-user/signal.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 89 insertions(+), 1 deletion(-)
bsd-user/main.c +5 -0
@@ -217,6 +217,11 @@ void qemu_cpu_kick(CPUState *cpu)/* Assumes contents are already zeroed. */ static void init_task_state (TaskState *ts) { + ts->sigaltstack_used = (struct target_sigaltstack ) { + .ss_sp =0 , + .ss_size =0 , + .ss_flags =TARGET_SS_DISABLE , + }; }void gemu_log (const char *fmt, ...)
bsd-user/qemu.h +2 -1
@@ -103,7 +103,8 @@ typedef struct TaskState {*/ sigset_t signal_mask; -uint8_t stack[]; +/* This thread's sigaltstack, if it has one */ +struct target_sigaltstack sigaltstack_used; } __attribute__((aligned(16))) TaskState;void stop_all_tasks (void );
bsd-user/signal.c +82 -0
94 lines not on this page · raw message
Signature · 1 line
2.33.1
PATCH v2 18/40 bsd-user/signal.c: Add si_type argument to queue_signal
In reply to Warner LoshMirror the linux-user practice and add a si_type argument to queue signal. This will be transported as the upper 8 bits in the si_type element of siginfo so that we know what bits of the structure are valid and so we can properly implement host_to_target_siginfo_noswap and tswap_siginfo. Adapt the one caller of queue_signal to the new interface. Use all the same names as Linux (except _RT which we don't treat differently, unlike Linux), though some are unused. Place this into signal-common.h since that's a better place given bsd-user's structure. Move prototype of queue_signal to signal-common.h to mirror linux-user's location.
--- bsd-user/signal-common.h | 26 +++++++++++++++++++++++++- bsd-user/signal.c | 5 +++-- 2 files changed, 28 insertions(+), 3 deletions(-)
bsd-user/signal-common.h +25 -1
@@ -15,8 +15,32 @@ long do_sigreturn(CPUArchState *env);void force_sig_fault (int sig,int code,abi_ulong addr);int host_to_target_signal (int sig);void process_pending_signals (CPUArchState *env); -void queue_signal (CPUArchState *env,int sig,target_siginfo_t *info); +void queue_signal (CPUArchState *env,int sig,int si_type, +target_siginfo_t *info);void signal_init (void );int target_to_host_signal (int sig); +/* +* Within QEMU the top 8 bits of si_code indicate which of the parts of the +* union in target_siginfo is valid. This only applies between +* host_to_target_siginfo_noswap() and tswap_siginfo(); it does not appear +* either within host siginfo_t or in target_siginfo structures which we get +* from the guest userspace program. Linux kenrels use this internally, but BSD +* kernels don't do this, but its a useful abstraction. +* +* The linux-user version of this uses the top 16 bits, but FreeBSD's SI_USER +* and other signal indepenent SI_ codes have bit 16 set, so we only use the top +* byte instead. +* +* For FreeBSD, we have si_pid, si_uid, si_status, and si_addr always. Linux and +* {Open,Net}BSD have a different approach (where their reason field is larger, +* but whose siginfo has fewer fields always). +*/ +#define QEMU_SI_NOINFO 0/* nothing other than si_signo valid */ +#define QEMU_SI_FAULT 1/* _fault is valid in _reason */ +#define QEMU_SI_TIMER 2/* _timer is valid in _reason */ +#define QEMU_SI_MESGQ 3/* _mesgq is valid in _reason */ +#define QEMU_SI_POLL 4/* _poll is valid in _reason */ +#define QEMU_SI_CAPSICUM 5/* _capsicum is valid in _reason */ +#endif
bsd-user/signal.c +3 -2
@@ -50,7 +50,8 @@ int target_to_host_signal(int sig)* Queue a signal so that it will be send to the virtual CPU as soon as * possible. */ -void queue_signal (CPUArchState *env,int sig,target_siginfo_t *info) +void queue_signal (CPUArchState *env,int sig,int si_type, +target_siginfo_t *info) {qemu_log_mask (LOG_UNIMP ,"No signal queueing, dropping signal %d\n" , sig); }
@@ -91,7 +92,7 @@ void force_sig_fault(int sig, int code, abi_ulong addr) info.si_errno =0 ; info.si_code = code; info.si_addr = addr; -queue_signal (env, sig, &info); +queue_signal (env, sig,QEMU_SI_FAULT , &info); }static void host_signal_handler (int host_sig,siginfo_t *info,void *puc)
Signature · 1 line
2.33.1
PATCH v2 15/40 bsd-user/arm/target_arch_cpu.h: Implement data faults
In reply to Warner LoshUpdate for the richer set of data faults that are now possible. Copied largely from linux-user/arm/cpu_loop.c, with minor typo fixes.
--- bsd-user/arm/target_arch_cpu.h | 45 ++++++++++++++++++++++++++-------- 1 file changed, 35 insertions(+), 10 deletions(-)
bsd-user/arm/target_arch_cpu.h +35 -10
@@ -39,8 +39,7 @@ static inline void target_cpu_init(CPUARMState *env,static inline void target_cpu_loop (CPUARMState *env) { -int trapnr; -target_siginfo_t info; +int trapnr, si_signo, si_code;unsigned int n;CPUState *cs =env_cpu (env);
@@ -150,15 +149,41 @@ static inline void target_cpu_loop(CPUARMState *env)/* just indicate that signals should be handled asap */ break ;case EXCP_PREFETCH_ABORT : -/* See arm/arm/trap.c prefetch_abort_handler() */ case EXCP_DATA_ABORT : -/* See arm/arm/trap.c data_abort_handler() */ - info.si_signo =TARGET_SIGSEGV ; - info.si_errno =0 ; -/* XXX: check env->error_code */ - info.si_code =0 ; - info.si_addr = env->exception.vaddress; -queue_signal (env, info.si_signo, &info); +/* +* See arm/arm/trap-v6.c prefetch_abort_handler() and +* data_abort_handler() +* +* However, FreeBSD maps these to a generic value and then uses that +* to maybe fault in pages in vm/vm_fault.c:vm_fault_trap(). I +* believe that the indirection maps the same as Linux, but haven't +* chased down every single possible indirection. +*/ + +/* For user-only we don't set TTBCR_EAE, so look at the FSR. */ +switch (env->exception.fsr &0x1f ) { +case 0x1 :/* Alignment */ + si_signo =TARGET_SIGBUS ; + si_code =TARGET_BUS_ADRALN ; +break ; +case 0x3 :/* Access flag fault, level 1 */ +case 0x6 :/* Access flag fault, level 2 */ +case 0x9 :/* Domain fault, level 1 */ +case 0xb :/* Domain fault, level 2 */ +case 0xd :/* Permission fault, level 1 */ +case 0xf :/* Permission fault, level 2 */ + si_signo =TARGET_SIGSEGV ; + si_code =TARGET_SEGV_ACCERR ; +break ; +case 0x5 :/* Translation fault, level 1 */ +case 0x7 :/* Translation fault, level 2 */ + si_signo =TARGET_SIGSEGV ; + si_code =TARGET_SEGV_MAPERR ; +break ; +default : +g_assert_not_reached (); + } +force_sig_fault (si_signo, si_code, env->exception.vaddress);break ;case EXCP_DEBUG :case EXCP_BKPT :
Signature · 1 line
2.33.1
PATCH v2 17/40 bsd-user/signal.c: Implement signal_init()
In reply to Warner LoshInitialize the signal state for the emulator. Setup a set of sane default signal handlers, mirroring the host's signals. For fatal signals (those that exit by default), establish our own set of signal handlers. Stub out the actual signal handler we use for the moment.
--- bsd-user/qemu.h | 7 +++++ bsd-user/signal.c | 67 +++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 74 insertions(+)
bsd-user/qemu.h +7 -0
@@ -94,6 +94,13 @@ typedef struct TaskState {* from multiple threads.) */ int signal_pending; +/* +* This thread's signal mask, as requested by the guest program. +* The actual signal mask of this thread may differ: +* + we don't let SIGSEGV and SIGBUS be blocked while running guest code +* + sometimes we block all signals to avoid races +*/ +sigset_t signal_mask;uint8_t stack[]; } __attribute__((aligned(16))) TaskState;
bsd-user/signal.c +67 -0
87 lines not on this page · raw message
Signature · 1 line
2.33.1
PATCH v2 34/40 bsd-user/signal.c: process_pending_signals
In reply to Warner LoshProcess the currently queued signals.
--- bsd-user/signal.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 55 insertions(+), 1 deletion(-)
bsd-user/signal.c +55 -1
@@ -755,8 +755,62 @@ static void handle_pending_signal(CPUArchState *env, int sig, } } -void process_pending_signals (CPUArchState *cpu_env) +void process_pending_signals (CPUArchState *env) { +CPUState *cpu =env_cpu (env); +int sig; +sigset_t *blocked_set, set; +struct emulated_sigtable *k; +TaskState *ts = cpu->opaque; + +while (qatomic_read (&ts->signal_pending)) { +sigfillset (&set); +sigprocmask (SIG_SETMASK , &set,0 ); + + restart_scan: + sig = ts->sync_signal.pending; +if (sig) { +/* +* Synchronous signals are forced by the emulated CPU in some way. +* If they are set to ignore, restore the default handler (see +* sys/kern_sig.c trapsignal() and execsigs() for this behavior) +* though maybe this is done only when forcing exit for non SIGCHLD. +*/ +if (sigismember (&ts->signal_mask,target_to_host_signal (sig)) || + sigact_table[sig -1 ]._sa_handler ==TARGET_SIG_IGN ) { +sigdelset (&ts->signal_mask,target_to_host_signal (sig)); + sigact_table[sig -1 ]._sa_handler =TARGET_SIG_DFL ; + } +handle_pending_signal (env, sig, &ts->sync_signal); + } + + k = ts->sigtab; +for (sig =1 ; sig <=TARGET_NSIG ; sig++, k++) { + blocked_set = ts->in_sigsuspend ? + &ts->sigsuspend_mask : &ts->signal_mask; +if (k->pending && + !sigismember (blocked_set,target_to_host_signal (sig))) { +handle_pending_signal (env, sig, k); +/* +* Restart scan from the beginning, as handle_pending_signal +* might have resulted in a new synchronous signal (eg SIGSEGV). +*/ + goto restart_scan; + } + } + +/* +* unblock signals and check one more time. Unblocking signals may cause +* us to take anothe rhost signal, which will set signal_pending again. +*/ +qatomic_set (&ts->signal_pending,0 ); + ts->in_sigsuspend = false; + set = ts->signal_mask; +sigdelset (&set,SIGSEGV ); +sigdelset (&set,SIGBUS ); +sigprocmask (SIG_SETMASK , &set,0 ); + } + ts->in_sigsuspend = false; } void cpu_loop_exit_sigsegv(CPUState *cpu, target_ulong addr,
Signature · 1 line
2.33.1
PATCH v2 23/40 bsd-user: Add trace events for bsd-user
In reply to Warner LoshAdd the bsd-user specific events and infrastructure. Only include the linux-user trace events for linux-user, not bsd-user.
--- bsd-user/signal.c | 1 + bsd-user/trace-events | 11 +++++++++++ bsd-user/trace.h | 1 + meson.build | 5 ++++- 4 files changed, 17 insertions(+), 1 deletion(-) create mode 100644 bsd-user/trace-events create mode 100644 bsd-user/trace.h
bsd-user/signal.c +1 -0
@@ -21,6 +21,7 @@#include "qemu/osdep.h" #include "qemu.h" #include "signal-common.h" +#include "trace.h" #include "hw/core/tcg-cpu-ops.h" #include "host-signal.h"
bsd-user/trace-events new file +11 -0
@@ -0,0 +1,11 @@ +# See docs/tracing.txt for syntax documentation. + +# bsd-user/signal.c +user_setup_frame(void *env, uint64_t frame_addr) "env=%p frame_addr=0x%"PRIx64 +user_setup_rt_frame(void *env, uint64_t frame_addr) "env=%p frame_addr=0x%"PRIx64 +user_do_rt_sigreturn(void *env, uint64_t frame_addr) "env=%p frame_addr=0x%"PRIx64 +user_do_sigreturn(void *env, uint64_t frame_addr) "env=%p frame_addr=0x%"PRIx64 +user_dump_core_and_abort(void *env, int target_sig, int host_sig) "env=%p signal %d (host %d)" +user_handle_signal(void *env, int target_sig) "env=%p signal %d" +user_host_signal(void *env, int host_sig, int target_sig) "env=%p signal %d (target %d(" +user_queue_signal(void *env, int target_sig) "env=%p signal %d"
bsd-user/trace.h new file +1 -0
@@ -0,0 +1 @@ +#include "trace/trace-bsd_user.h"
meson.build +4 -1
@@ -2437,9 +2437,12 @@ trace_events_subdirs = [ 'monitor', 'util', ] -if have_user +if have_linux_user trace_events_subdirs += [ 'linux-user' ] endif +if have_bsd_user + trace_events_subdirs += [ 'bsd-user' ] +endif if have_block trace_events_subdirs += [ 'authz',
Signature · 1 line
2.33.1
PATCH v2 16/40 bsd-user/signal.c: implement abstract target / host signal translation
In reply to Warner LoshImplement host_to_target_signal and target_to_host_signal.
--- bsd-user/signal-common.h | 2 ++ bsd-user/signal.c | 16 ++++++++++++++++ 2 files changed, 18 insertions(+)
bsd-user/signal-common.h +2 -0
@@ -13,8 +13,10 @@ long do_rt_sigreturn(CPUArchState *env);abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp);long do_sigreturn (CPUArchState *env);void force_sig_fault (int sig,int code,abi_ulong addr); +int host_to_target_signal (int sig);void process_pending_signals (CPUArchState *env);void queue_signal (CPUArchState *env,int sig,target_siginfo_t *info);void signal_init (void ); +int target_to_host_signal (int sig);#endif
bsd-user/signal.c +16 -0
@@ -2,6 +2,7 @@* Emulation of BSD signals * * Copyright (c) 2003 - 2008 Fabrice Bellard +* Copyright (c) 2013 Stacey Son * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by
@@ -27,6 +28,21 @@* fork. */ +/* +* The BSD ABIs use the same singal numbers across all the CPU architectures, so +* (unlike Linux) these functions are just the identity mapping. This might not +* be true for XyzBSD running on AbcBSD, which doesn't currently work. +*/ +int host_to_target_signal (int sig) +{ +return sig; +} + +int target_to_host_signal (int sig) +{ +return sig; +} +/* * Queue a signal so that it will be send to the virtual CPU as soon as * possible.
Signature · 1 line
2.33.1
PATCH v2 19/40 bsd-user/host/arm/host-signal.h: Implement host_signal_*
In reply to Warner LoshImplement host_signal_pc, host_signal_set_pc and host_signal_write for arm.
--- bsd-user/host/arm/host-signal.h | 35 +++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+) create mode 100644 bsd-user/host/arm/host-signal.h
bsd-user/host/arm/host-signal.h new file +35 -0
@@ -0,0 +1,35 @@ +/* +* host-signal.h: signal info dependent on the host architecture +* +* Copyright (c) 2021 Warner Losh +* +* SPDX-License-Identifier: GPL-2.0-or-later +*/ + +#ifndef ARM_HOST_SIGNAL_H +#define ARM_HOST_SIGNAL_H + +#include <sys/ucontext.h> + +static inline uintptr_t host_signal_pc (ucontext_t *uc) +{ +return uc->uc_mcontext.__gregs[_REG_PC]; +} + +static inline void host_signal_set_pc (ucontext_t *uc,uintptr_t pc) +{ + uc->uc_mcontext.__gregs[_REG_PC] = pc; +} + +static inline bool host_signal_write (siginfo_t *info,ucontext_t *uc) +{ +/* +* In the FSR, bit 11 is WnR. FreeBSD returns this as part of the +* si_info.si_trapno. +*/ +uint32_t fsr = info->si_trapno; + +return extract32 (fsr,11 ,1 ); +} + +#endif
Signature · 1 line
2.33.1
PATCH v2 39/40 bsd-user: Rename arg name for target_cpu_reset to env
In reply to Warner LoshRename the parameter name for target_cpu_reset's CPUArchState * arg from cpu to env.
--- bsd-user/arm/target_arch_cpu.h | 2 +- bsd-user/i386/target_arch_cpu.h | 4 ++-- bsd-user/x86_64/target_arch_cpu.h | 4 ++-- 3 files changed, 5 insertions(+), 5 deletions(-)
bsd-user/arm/target_arch_cpu.h +1 -1
@@ -213,7 +213,7 @@ static inline void target_cpu_clone_regs(CPUARMState *env, target_ulong newsp) env->regs[0 ] =0 ; } -static inline void target_cpu_reset (CPUArchState *cpu) +static inline void target_cpu_reset (CPUArchState *env) { }
bsd-user/i386/target_arch_cpu.h +2 -2
@@ -199,9 +199,9 @@ static inline void target_cpu_clone_regs(CPUX86State *env, target_ulong newsp) env->regs[R_EAX] = 0; } -static inline void target_cpu_reset (CPUArchState *cpu) +static inline void target_cpu_reset (CPUArchState *env) { -cpu_reset (env_cpu (cpu)); +cpu_reset (env_cpu (env)); }#endif /* ! _TARGET_ARCH_CPU_H_ */
bsd-user/x86_64/target_arch_cpu.h +2 -2
@@ -237,9 +237,9 @@ static inline void target_cpu_clone_regs(CPUX86State *env, target_ulong newsp) env->regs[R_EAX] = 0; } -static inline void target_cpu_reset (CPUArchState *cpu) +static inline void target_cpu_reset (CPUArchState *env) { -cpu_reset (env_cpu (cpu)); +cpu_reset (env_cpu (env)); }#endif /* ! _TARGET_ARCH_CPU_H_ */
Signature · 1 line
2.33.1
PATCH v2 26/40 bsd-user/signal.c: Implement host_signal_handler
In reply to Warner LoshImplement host_signal_handler to handle signals generated by the host and to do safe system calls.
--- bsd-user/signal.c | 105 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 105 insertions(+)
bsd-user/signal.c +105 -0
@@ -223,6 +223,111 @@ void force_sig_fault(int sig, int code, abi_ulong addr)static void host_signal_handler (int host_sig,siginfo_t *info,void *puc) { +CPUArchState *env = thread_cpu->env_ptr; +CPUState *cpu =env_cpu (env); +TaskState *ts = cpu->opaque; +target_siginfo_t tinfo; +ucontext_t *uc = puc; +struct emulated_sigtable *k; +int guest_sig; +uintptr_t pc =0 ; +bool sync_sig = false; + +/* +* Non-spoofed SIGSEGV and SIGBUS are synchronous, and need special +* handling wrt signal blocking and unwinding. +*/ +if ((host_sig ==SIGSEGV || host_sig ==SIGBUS ) && info->si_code >0 ) { +MMUAccessType access_type; +uintptr_t host_addr; +abi_ptr guest_addr; +bool is_write; + + host_addr = (uintptr_t )info->si_addr; + +/* +* Convert forcefully to guest address space: addresses outside +* reserved_va are still valid to report via SEGV_MAPERR. +*/ + guest_addr =h2g_nocheck (host_addr); + + pc =host_signal_pc (uc); + is_write =host_signal_write (info, uc); + access_type =adjust_signal_pc (&pc, is_write); + +if (host_sig ==SIGSEGV ) { +bool maperr = true; + +if (info->si_code ==SEGV_ACCERR &&h2g_valid (host_addr)) { +/* If this was a write to a TB protected page, restart. */ +if (is_write && +handle_sigsegv_accerr_write (cpu, &uc->uc_sigmask, + pc, guest_addr)) { +return ; + } + +/* +* With reserved_va, the whole address space is PROT_NONE, +* which means that we may get ACCERR when we want MAPERR. +*/ +if (page_get_flags (guest_addr) &PAGE_VALID ) { + maperr = false; + }else { + info->si_code =SEGV_MAPERR ; + } + } + +sigprocmask (SIG_SETMASK , &uc->uc_sigmask,NULL ); +cpu_loop_exit_sigsegv (cpu, guest_addr, access_type, maperr, pc); + }else { +sigprocmask (SIG_SETMASK , &uc->uc_sigmask,NULL ); +if (info->si_code ==BUS_ADRALN ) { +cpu_loop_exit_sigbus (cpu, guest_addr, access_type, pc); + } + } + + sync_sig = true; + } + +/* Get the target signal number. */ + guest_sig =host_to_target_signal (host_sig); +if (guest_sig <1 || guest_sig >TARGET_NSIG ) { +return ; + } +trace_user_host_signal (cpu, host_sig, guest_sig); + +host_to_target_siginfo_noswap (&tinfo, info); + + k = &ts->sigtab[guest_sig -1 ]; + k->info = tinfo; + k->pending = guest_sig; + ts->signal_pending =1 ; + +/* +* For synchronous signals, unwind the cpu state to the faulting +* insn and then exit back to the main loop so that the signal +* is delivered immediately. +*/ +if (sync_sig) { + cpu->exception_index =EXCP_INTERRUPT ; +cpu_loop_exit_restore (cpu, pc); + } + +rewind_if_in_safe_syscall (puc); + +/* +* Block host signals until target signal handler entered. We +* can't block SIGSEGV or SIGBUS while we're executing guest +* code in case the guest code provokes one in the window between +* now and it getting out to the main loop. Signals will be +* unblocked again in process_pending_signals(). +*/ +sigfillset (&uc->uc_sigmask); +sigdelset (&uc->uc_sigmask,SIGSEGV ); +sigdelset (&uc->uc_sigmask,SIGBUS ); + +/* Interrupt the virtual CPU as soon as possible. */ +cpu_exit (thread_cpu); }void signal_init (void )
Signature · 1 line
2.33.1
PATCH v2 30/40 bsd-user/signal.c: sigset manipulation routines.
In reply to Warner Loshtarget_sigemptyset: resets a set to having no bits settarget_sigaddset: adds a signal to a settarget_sigismember: returns true when signal is a member host_to_target_sigset_internal: convert host sigset to targethost_to_target_sigset: convert host sigset to target target_to_host_sigset_internal: convert target sigset to hosttarget_to_host_sigset: convert target sigset to host
--- bsd-user/signal-common.h | 2 ++ bsd-user/signal.c | 74 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 76 insertions(+)
bsd-user/signal-common.h +2 -0
@@ -14,11 +14,13 @@ abi_long do_sigaltstack(abi_ulong uss_addr, abi_ulong uoss_addr, abi_ulong sp);long do_sigreturn (CPUArchState *env);void force_sig_fault (int sig,int code,abi_ulong addr);int host_to_target_signal (int sig); +void host_to_target_sigset (target_sigset_t *d,const sigset_t *s);void process_pending_signals (CPUArchState *env);void queue_signal (CPUArchState *env,int sig,int si_type,target_siginfo_t *info);void signal_init (void );int target_to_host_signal (int sig); +void target_to_host_sigset (sigset_t *d,const target_sigset_t *s);/* * Within QEMU the top 8 bits of si_code indicate which of the parts of the
bsd-user/signal.c +74 -0
92 lines not on this page · raw message
Signature · 1 line
2.33.1
PATCH v2 27/40 bsd-user/strace.c: print_taken_signal
In reply to Warner Loshprint_taken_signal() prints signals when we're tracing signals.
--- bsd-user/qemu.h | 10 +++++ bsd-user/strace.c | 97 +++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 107 insertions(+)
bsd-user/qemu.h +10 -0
@@ -204,6 +204,16 @@ print_openbsd_syscall(int num, abi_long arg1, abi_long arg2, abi_long arg3, abi_long arg4, abi_long arg5, abi_long arg6);void print_openbsd_syscall_ret (int num,abi_long ret); +/** +* print_taken_signal: +* @target_signum: target signal being taken +* @tinfo: target_siginfo_t which will be passed to the guest for the signal +* +* Print strace output indicating that this signal is being taken by the guest, +* in a format similar to: +* --- SIGSEGV {si_signo=SIGSEGV, si_code=SI_KERNEL, si_addr=0} --- +*/ +void print_taken_signal (int target_signum,const target_siginfo_t *tinfo);extern int do_strace;/* mmap.c */
bsd-user/strace.c +97 -0
106 lines not on this page · raw message
Signature · 1 line
2.33.1
PATCH v2 21/40 bsd-user/host/x86_64/host-signal.h: Implement host_signal_*
In reply to Warner LoshImplement host_signal_pc, host_signal_set_pc and host_signal_write for x86_64.
--- bsd-user/host/x86_64/host-signal.h | 37 ++++++++++++++++++++++++++++++ 1 file changed, 37 insertions(+) create mode 100644 bsd-user/host/x86_64/host-signal.h
bsd-user/host/x86_64/host-signal.h new file +37 -0
@@ -0,0 +1,37 @@ +/* +* host-signal.h: signal info dependent on the host architecture +* +* Copyright (c) 2021 Warner Losh +* +* SPDX-License-Identifier: GPL-2.0-or-later +*/ + +#ifndef X86_64_HOST_SIGNAL_H +#define X86_64_HOST_SIGNAL_H + +#include <sys/ucontext.h> +#include <machine/trap.h> +#include <vm/pmap.h> +#include <machine/pmap.h> + +static inline uintptr_t host_signal_pc (ucontext_t *uc) +{ +return uc->uc_mcontext.mc_rip; +} + +static inline void host_signal_set_pc (ucontext_t *uc,uintptr_t pc) +{ + uc->uc_mcontext.mc_rip = pc; +} + +static inline bool host_signal_write (siginfo_t *info,ucontext_t *uc) +{ +/* +* Look in sys/amd64/amd64/trap.c. NOTE: mc_err == tr_err due to type +* punning between a trapframe and mcontext on FreeBSD/amd64. +*/ +return uc->uc_mcontext.mc_trapno ==T_PAGEFLT && + uc->uc_mcontext.mc_err &PGEX_W ; +} + +#endif
Signature · 1 line
2.33.1
PATCH v2 40/40 bsd-user/freebsd/target_os_ucontext.h: Prefer env as arg name for CPUArchState args
In reply to Warner Losh--- bsd-user/freebsd/target_os_ucontext.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
bsd-user/freebsd/target_os_ucontext.h +3 -3
@@ -36,9 +36,9 @@ abi_long set_sigtramp_args(CPUArchState *env, int sig,struct target_sigframe *frame, abi_ulong frame_addr, struct target_sigaction *ka); -abi_long get_mcontext (CPUArchState *regs,target_mcontext_t *mcp,int flags); -abi_long set_mcontext (CPUArchState *regs,target_mcontext_t *mcp,int srflag); -abi_long get_ucontext_sigreturn (CPUArchState *regs,abi_ulong target_sf, +abi_long get_mcontext (CPUArchState *env,target_mcontext_t *mcp,int flags); +abi_long set_mcontext (CPUArchState *env,target_mcontext_t *mcp,int srflag); +abi_long get_ucontext_sigreturn (CPUArchState *env,abi_ulong target_sf,abi_ulong *target_uc);#endif /* TARGET_OS_UCONTEXT_H */
Signature · 1 line
2.33.1
PATCH v2 36/40 bsd-user/signal.c: implement do_sigaction
In reply to Warner LoshImplement the meat of the sigaction(2) system call with do_sigaction and helper routiner block_signals (which is also used to implemement signal masking so it's global).
Pending Comments from Peter Maydell <[email protected]>
(1) in block_signals, sigprocmast
For linux-user we rely on sigprocmask() in a multithreaded program setting the signal mask for only the calling thread, which isn't POSIX-mandated. (Arguably we should use pthread_sigmask() instead, but we don't for basically historical reasons since linux-user is host-OS-specific anyway.) Does BSD have the same "this changes this thread's signal mask" semantics for sigprocmask()?
(2) do_sigaction, first if validating which signals can be sent Kernel seems to allow SIGKILL and SIGSTOP unless act is non-NULL and act->sa_handler is SIG_DFL ?
https://github.com/freebsd/freebsd-src/blob/main/sys/kern/kern_sig.c#L747
(Compare linux-user commit ee3500d33a7431, a recent bugfix.)
(3) Noting confusion in do_sigaction between host and target errnos (they are identical on BSD, but we should still return the right sort in case it ever does matter).
--- bsd-user/signal-common.h | 22 ++++++++++++ bsd-user/signal.c | 76 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 98 insertions(+)
bsd-user/signal-common.h +22 -0
@@ -9,7 +9,29 @@#ifndef SIGNAL_COMMON_H #define SIGNAL_COMMON_H +/** +* block_signals: block all signals while handling this guest syscall +* +* Block all signals, and arrange that the signal mask is returned to +* its correct value for the guest before we resume execution of guest code. +* If this function returns non-zero, then the caller should immediately +* return -TARGET_ERESTARTSYS to the main loop, which will take the pending +* signal and restart execution of the syscall. +* If block_signals() returns zero, then the caller can continue with +* emulation of the system call knowing that no signals can be taken +* (and therefore that no race conditions will result). +* This should only be called once, because if it is called a second time +* it will always return non-zero. (Think of it like a mutex that can't +* be recursively locked.) +* Signals will be unblocked again by process_pending_signals(). +* +* Return value: non-zero if there was a pending signal, zero if not. +*/ +int block_signals (void );/* Returns non zero if signal pending */ +long do_rt_sigreturn (CPUArchState *env); +int do_sigaction (int sig,const struct target_sigaction *act, +struct target_sigaction *oact);abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp);long do_sigreturn (CPUArchState *env,abi_ulong addr);void force_sig_fault (int sig,int code,abi_ulong addr);
bsd-user/signal.c +76 -0
88 lines not on this page · raw message
Signature · 1 line
2.33.1
PATCH v2 24/40 bsd-user/signal.c: host_to_target_siginfo_noswap
In reply to Warner LoshImplement conversion of host to target siginfo.
--- bsd-user/signal.c | 113 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 113 insertions(+)
bsd-user/signal.c +113 -0
@@ -48,6 +48,119 @@ int target_to_host_signal(int sig)return sig; } +static bool has_trapno (int tsig) +{ +return tsig ==TARGET_SIGILL || + tsig ==TARGET_SIGFPE || + tsig ==TARGET_SIGSEGV || + tsig ==TARGET_SIGBUS || + tsig ==TARGET_SIGTRAP ; +} + + +/* Siginfo conversion. */ + +/* +* Populate tinfo w/o swapping based on guessing which fields are valid. +*/ +static inline void host_to_target_siginfo_noswap (target_siginfo_t *tinfo, +const siginfo_t *info) +{ +int sig =host_to_target_signal (info->si_signo); +int si_code = info->si_code; +int si_type; + +/* +* Make sure we that the variable portion of the target siginfo is zeroed +* out so we don't leak anything into that. +*/ +memset (&tinfo->_reason,0 ,sizeof (tinfo->_reason)); + +/* +* This is awkward, because we have to use a combination of the si_code and +* si_signo to figure out which of the union's members are valid.o We +* therefore make our best guess. +* +* Once we have made our guess, we record it in the top 16 bits of +* the si_code, so that tswap_siginfo() later can use it. +* tswap_siginfo() will strip these top bits out before writing +* si_code to the guest (sign-extending the lower bits). +*/ + tinfo->si_signo = sig; + tinfo->si_errno = info->si_errno; + tinfo->si_code = info->si_code; + tinfo->si_pid = info->si_pid; + tinfo->si_uid = info->si_uid; + tinfo->si_status = info->si_status; + tinfo->si_addr = (abi_ulong )(unsigned long )info->si_addr; +/* +* si_value is opaque to kernel. On all FreeBSD platforms, +* sizeof(sival_ptr) >= sizeof(sival_int) so the following +* always will copy the larger element. +*/ + tinfo->si_value.sival_ptr = + (abi_ulong )(unsigned long )info->si_value.sival_ptr; + +switch (si_code) { +/* +* All the SI_xxx codes that are defined here are global to +* all the signals (they have values that none of the other, +* more specific signal info will set). +*/ +case SI_USER : +case SI_LWP : +case SI_KERNEL : +case SI_QUEUE : +case SI_ASYNCIO : +/* +* Only the fixed parts are valid (though FreeBSD doesn't always +* set all the fields to non-zero values. +*/ + si_type =QEMU_SI_NOINFO ; +break ; +case SI_TIMER : + tinfo->_reason._timer._timerid = info->_reason._timer._timerid; + tinfo->_reason._timer._overrun = info->_reason._timer._overrun; + si_type =QEMU_SI_TIMER ; +break ; +case SI_MESGQ : + tinfo->_reason._mesgq._mqd = info->_reason._mesgq._mqd; + si_type =QEMU_SI_MESGQ ; +break ; +default : +/* +* We have to go based on the signal number now to figure out +* what's valid. +*/ +if (has_trapno (sig)) { + tinfo->_reason._fault._trapno = info->_reason._fault._trapno; + si_type =QEMU_SI_FAULT ; + } +#ifdef TARGET_SIGPOLL +/* +* FreeBSD never had SIGPOLL, but emulates it for Linux so there's +* a chance it may popup in the future. +*/ +if (sig ==TARGET_SIGPOLL ) { + tinfo->_reason._poll._band = info->_reason._poll._band; + si_type =QEMU_SI_POLL ; + } +#endif +/* +* Unsure that this can actually be generated, and our support for +* capsicum is somewhere between weak and non-existant, but if we get +* one, then we know what to save. +*/ +if (sig ==TARGET_SIGTRAP ) { + tinfo->_reason._capsicum._syscall = + info->_reason._capsicum._syscall; + si_type =QEMU_SI_CAPSICUM ; + } +break ; + } + tinfo->si_code =deposit32 (si_code,24 ,8 , si_type); +} +/* * Queue a signal so that it will be send to the virtual CPU as soon as * possible.
Signature · 1 line
2.33.1
PATCH v2 25/40 bsd-user/signal.c: Implement rewind_if_in_safe_syscall
In reply to Warner Losh--- bsd-user/qemu.h | 2 ++ bsd-user/signal.c | 13 ++++++++++++- 2 files changed, 14 insertions(+), 1 deletion(-)
bsd-user/qemu.h +2 -0
@@ -446,4 +446,6 @@ static inline void *lock_user_string(abi_ulong guest_addr)#include <pthread.h> +#include "user/safe-syscall.h" +#endif /* QEMU_H */
bsd-user/signal.c +12 -1
@@ -48,6 +48,18 @@ int target_to_host_signal(int sig)return sig; } +/* Adjust the signal context to rewind out of safe-syscall if we're in it */ +static inline void rewind_if_in_safe_syscall (void *puc) +{ +ucontext_t *uc = (ucontext_t *)puc; +uintptr_t pcreg =host_signal_pc (uc); + +if (pcreg > (uintptr_t )safe_syscall_start + && pcreg < (uintptr_t )safe_syscall_end) { +host_signal_set_pc (uc, (uintptr_t )safe_syscall_start); + } +} +static bool has_trapno (int tsig) { return tsig == TARGET_SIGILL ||
@@ -57,7 +69,6 @@ static bool has_trapno(int tsig) tsig ==TARGET_SIGTRAP ; } -/* Siginfo conversion. */ /*
Signature · 1 line
2.33.1
PATCH v2 38/40 MAINTAINERS: Add tests/vm/*bsd to the list to get reviews on
In reply to Warner Loshtests/vm/*bsd (especailly tests/vm/freebsd) are adjacent to the bsd-user stuff and we're keen on keeping them working as well.
--- MAINTAINERS | 1 + 1 file changed, 1 insertion(+)
MAINTAINERS +1 -0
@@ -3101,6 +3101,7 @@ R: Kyle Evans <[email protected]> S: Maintained F: bsd-user/ F: configs/targets/*-bsd-user.mak +F: tests/vm/*bsd T: git https://github.com/qemu-bsd-user/qemu-bsd-user bsd-user-rebase-3.1 Linux user
Signature · 1 line
2.33.1
PATCH v2 29/40 bsd-user/signal.c: Fill in queue_signal
In reply to Warner LoshFill in queue signal implementation, as well as routines allocate and delete elements of the signal queue.
--- bsd-user/qemu.h | 1 + bsd-user/signal.c | 13 ++++++++++++- 2 files changed, 13 insertions(+), 1 deletion(-)
bsd-user/qemu.h +1 -0
@@ -85,6 +85,7 @@ typedef struct TaskState {struct bsd_binprm *bprm;struct image_info *info; +struct emulated_sigtable sync_signal;struct emulated_sigtable sigtab[TARGET_NSIG ];/* * Nonzero if process_pending_signals() needs to do something (either
bsd-user/signal.c +12 -1
@@ -255,7 +255,18 @@ static void QEMU_NORETURN dump_core_and_abort(int target_sig)void queue_signal (CPUArchState *env,int sig,int si_type,target_siginfo_t *info) { -qemu_log_mask (LOG_UNIMP ,"No signal queueing, dropping signal %d\n" , sig); +CPUState *cpu =env_cpu (env); +TaskState *ts = cpu->opaque; + +trace_user_queue_signal (env, sig); + + info->si_code =deposit32 (info->si_code,24 ,8 , si_type); + + ts->sync_signal.info = *info; + ts->sync_signal.pending = sig; +/* Signal that a new signal is pending. */ +qatomic_set (&ts->signal_pending,1 ); +return ; }static int fatal_signal (int sig)
Signature · 1 line
2.33.1
PATCH v2 22/40 bsd-user: Add host signals to the build
In reply to Warner LoshStart to add the host signal functionality to the build.
--- bsd-user/meson.build | 1 + bsd-user/signal.c | 1 + meson.build | 1 + 3 files changed, 3 insertions(+)
bsd-user/meson.build +1 -0
@@ -4,6 +4,7 @@ endif bsd_user_ss = ss.source_set() +common_user_inc += include_directories('host/' / host_arch) common_user_inc += include_directories('.') bsd_user_ss.add(files(
bsd-user/signal.c +1 -0
@@ -22,6 +22,7 @@#include "qemu.h" #include "signal-common.h" #include "hw/core/tcg-cpu-ops.h" +#include "host-signal.h" /* * Stubbed out routines until we merge signal support from bsd-user
meson.build +1 -0
@@ -2932,6 +2932,7 @@ foreach target : target_dirs if 'CONFIG_BSD_USER' in config_target base_dir = 'bsd-user' target_inc += include_directories('bsd-user/' / targetos) + target_inc += include_directories('bsd-user/host/' / host_arch) dir = base_dir / abi arch_srcs += files(dir / 'signal.c', dir / 'target_arch_cpu.c') endif
Signature · 1 line
2.33.1
PATCH v2 28/40 bsd-user/signal.c: Implement dump_core_and_abort
In reply to Warner LoshForce delivering a signal and generating a core file. It's a global function for the moment...
--- bsd-user/signal.c | 76 +++++++++++++++++++++++++++++++++++++++++ bsd-user/syscall_defs.h | 1 + 2 files changed, 77 insertions(+)
bsd-user/signal.c +76 -0
@@ -172,6 +172,82 @@ static inline void host_to_target_siginfo_noswap(target_siginfo_t *tinfo, tinfo->si_code =deposit32 (si_code,24 ,8 , si_type); } +/* Returns 1 if given signal should dump core if not handled. */ +static int core_dump_signal (int sig) +{ +switch (sig) { +case TARGET_SIGABRT : +case TARGET_SIGFPE : +case TARGET_SIGILL : +case TARGET_SIGQUIT : +case TARGET_SIGSEGV : +case TARGET_SIGTRAP : +case TARGET_SIGBUS : +return 1 ; +default : +return 0 ; + } +} + +/* Abort execution with signal. */ +static void QEMU_NORETURN dump_core_and_abort(int target_sig) +{ +CPUArchState *env = thread_cpu->env_ptr; +CPUState *cpu =env_cpu (env); +TaskState *ts = cpu->opaque; +int core_dumped =0 ; +int host_sig; +struct sigaction act; + + host_sig =target_to_host_signal (target_sig); +gdb_signalled (env, target_sig); + +/* Dump core if supported by target binary format */ +if (core_dump_signal (target_sig) && (ts->bprm->core_dump !=NULL )) { +stop_all_tasks (); + core_dumped = + ((*ts->bprm->core_dump)(target_sig, env) ==0 ); + } +if (core_dumped) { +struct rlimit nodump; + +/* +* We already dumped the core of target process, we don't want +* a coredump of qemu itself. +*/ +getrlimit (RLIMIT_CORE , &nodump); + nodump.rlim_cur =0 ; +setrlimit (RLIMIT_CORE , &nodump); + (void )fprintf (stderr,"qemu: uncaught target signal %d (%s) " +"- %s\n" , target_sig,strsignal (host_sig),"core dumped" ); + } + +/* +* The proper exit code for dying from an uncaught signal is +* -<signal>. The kernel doesn't allow exit() or _exit() to pass +* a negative value. To get the proper exit code we need to +* actually die from an uncaught signal. Here the default signal +* handler is installed, we send ourself a signal and we wait for +* it to arrive. +*/ +memset (&act,0 ,sizeof (act)); +sigfillset (&act.sa_mask); + act.sa_handler =SIG_DFL ; +sigaction (host_sig, &act,NULL ); + +kill (getpid (), host_sig); + +/* +* Make sure the signal isn't masked (just reuse the mask inside +* of act). +*/ +sigdelset (&act.sa_mask, host_sig); +sigsuspend (&act.sa_mask); + +/* unreachable */ +abort (); +} +/* * Queue a signal so that it will be send to the virtual CPU as soon as * possible.
bsd-user/syscall_defs.h +1 -0
@@ -21,6 +21,7 @@#define _SYSCALL_DEFS_H_#include <sys/syscall.h> +#include <sys/resource.h> #include "errno_defs.h"
Signature · 1 line
2.33.1
PATCH v2 32/40 bsd-user/signal.c: handle_pending_signal
In reply to Warner LoshHandle a queued signal.
--- bsd-user/qemu.h | 7 ++++ bsd-user/signal.c | 87 +++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 94 insertions(+)
bsd-user/qemu.h +7 -0
@@ -95,6 +95,8 @@ typedef struct TaskState {* from multiple threads.) */ int signal_pending; +/* True if we're leaving a sigsuspend and sigsuspend_mask is valid. */ +bool in_sigsuspend;/* * This thread's signal mask, as requested by the guest program. * The actual signal mask of this thread may differ:
@@ -102,6 +104,11 @@ typedef struct TaskState {* + sometimes we block all signals to avoid races */ sigset_t signal_mask; +/* +* The signal mask imposed by a guest sigsuspend syscall, if we are +* currently in the middle of such a syscall +*/ +sigset_t sigsuspend_mask;/* This thread's sigaltstack, if it has one */ struct target_sigaltstack sigaltstack_used;
bsd-user/signal.c +87 -0
93 lines not on this page · raw message
Signature · 1 line
2.33.1
PATCH v2 35/40 bsd-user/signal.c: implement do_sigreturn
In reply to Warner LoshImplements the meat of a sigreturn(2) system call via do_sigreturn, and helper reset_signal_mask. Fix the prototype of do_sigreturn in qemu.h and remove do_rt_sigreturn since it's linux only.
PENDING COMMENTS: from Peter Maydell <[email protected]>
do_sigreturn() itself shouldn't be setting the active signal mask, at least if you follow the linux-user design. It just sets the thread's signal_mask field in the TaskState by calling set_sigmask(), and then on our way out in the main cpu loop we'll call process_pending_signals() which sets the real thread signal mask to that value. (This, together with do_sigreturn() calling block_signals() before it starts work, avoids some race conditions where a host signal is delivered as soon as we unblock, I think.)
--- bsd-user/signal-common.h | 2 +- bsd-user/signal.c | 56 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 57 insertions(+), 1 deletion(-)
bsd-user/signal-common.h +1 -1
@@ -11,7 +11,7 @@long do_rt_sigreturn (CPUArchState *env);abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp); -long do_sigreturn (CPUArchState *env); +long do_sigreturn (CPUArchState *env,abi_ulong addr);void force_sig_fault (int sig,int code,abi_ulong addr);int host_to_target_signal (int sig);void host_to_target_sigset (target_sigset_t *d,const sigset_t *s);
bsd-user/signal.c +56 -0
62 lines not on this page · raw message
Signature · 1 line
2.33.1
PATCH v2 33/40 bsd-user/signal.c: tswap_siginfo
In reply to Warner LoshConvert siginfo from targer to host.
--- bsd-user/signal.c | 53 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 53 insertions(+)
bsd-user/signal.c +53 -0
@@ -256,6 +256,59 @@ static inline void host_to_target_siginfo_noswap(target_siginfo_t *tinfo, tinfo->si_code =deposit32 (si_code,24 ,8 , si_type); } +static void tswap_siginfo (target_siginfo_t *tinfo,const target_siginfo_t *info) +{ +int si_type =extract32 (info->si_code,24 ,8 ); +int si_code =sextract32 (info->si_code,0 ,24 ); + +__put_user (info->si_signo, &tinfo->si_signo); +__put_user (info->si_errno, &tinfo->si_errno); +__put_user (si_code, &tinfo->si_code);/* Zero out si_type, it's internal */ +__put_user (info->si_pid, &tinfo->si_pid); +__put_user (info->si_uid, &tinfo->si_uid); +__put_user (info->si_status, &tinfo->si_status); +__put_user (info->si_addr, &tinfo->si_addr); +/* +* Unswapped, because we passed it through mostly untouched. si_value is +* opaque to the kernel, so we didn't bother with potentially wasting cycles +* to swap it into host byte order. +*/ + tinfo->si_value.sival_ptr = info->si_value.sival_ptr; + +/* +* We can use our internal marker of which fields in the structure +* are valid, rather than duplicating the guesswork of +* host_to_target_siginfo_noswap() here. +*/ +switch (si_type) { +case QEMU_SI_NOINFO :/* No additional info */ +break ; +case QEMU_SI_FAULT : +__put_user (info->_reason._fault._trapno, + &tinfo->_reason._fault._trapno); +break ; +case QEMU_SI_TIMER : +__put_user (info->_reason._timer._timerid, + &tinfo->_reason._timer._timerid); +__put_user (info->_reason._timer._overrun, + &tinfo->_reason._timer._overrun); +break ; +case QEMU_SI_MESGQ : +__put_user (info->_reason._mesgq._mqd, &tinfo->_reason._mesgq._mqd); +break ; +case QEMU_SI_POLL : +/* Note: Not generated on FreeBSD */ +__put_user (info->_reason._poll._band, &tinfo->_reason._poll._band); +break ; +case QEMU_SI_CAPSICUM : +__put_user (info->_reason._capsicum._syscall, + &tinfo->_reason._capsicum._syscall); +break ; +default : +g_assert_not_reached (); + } +} +/* Returns 1 if given signal should dump core if not handled. */ static int core_dump_signal (int sig) {
Signature · 1 line
2.33.1
PATCH v2 37/40 bsd-user/signal.c: do_sigaltstack
In reply to Warner LoshImplement the meat of the sigaltstack(2) system call with do_sigaltstack.
With that, all the stubbed out routines are complete, so remove now-incorrect comment.
--- bsd-user/signal.c | 72 +++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 67 insertions(+), 5 deletions(-)
bsd-user/signal.c +67 -5
@@ -25,11 +25,6 @@#include "hw/core/tcg-cpu-ops.h" #include "host-signal.h" -/* -* Stubbed out routines until we merge signal support from bsd-user -* fork. -*/ -static struct target_sigaction sigact_table[TARGET_NSIG ];static void host_signal_handler (int host_sig,siginfo_t *info,void *puc); static void target_to_host_sigset_internal(sigset_t *d,
@@ -570,6 +565,73 @@ static void host_signal_handler(int host_sig, siginfo_t *info, void *puc)cpu_exit (thread_cpu); } +/* do_sigaltstack() returns target values and errnos. */ +/* compare to kern/kern_sig.c sys_sigaltstack() and kern_sigaltstack() */ +abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp) +{ +TaskState *ts = (TaskState *)thread_cpu->opaque; +int ret; +target_stack_t oss; + +if (uoss_addr) { +/* Save current signal stack params */ + oss.ss_sp =tswapl (ts->sigaltstack_used.ss_sp); + oss.ss_size =tswapl (ts->sigaltstack_used.ss_size); + oss.ss_flags =tswapl (sas_ss_flags (ts, sp)); + } + +if (uss_addr) { +target_stack_t *uss; +target_stack_t ss; +size_t minstacksize =TARGET_MINSIGSTKSZ ; + + ret = -TARGET_EFAULT ; +if (!lock_user_struct (VERIFY_READ , uss, uss_addr,1 )) { + goto out; + } +__get_user (ss.ss_sp, &uss->ss_sp); +__get_user (ss.ss_size, &uss->ss_size); +__get_user (ss.ss_flags, &uss->ss_flags); +unlock_user_struct (uss, uss_addr,0 ); + + ret = -TARGET_EPERM ; +if (on_sig_stack (ts, sp)) { + goto out; + } + + ret = -TARGET_EINVAL ; +if (ss.ss_flags !=TARGET_SS_DISABLE + && ss.ss_flags !=TARGET_SS_ONSTACK + && ss.ss_flags !=0 ) { + goto out; + } + +if (ss.ss_flags ==TARGET_SS_DISABLE ) { + ss.ss_size =0 ; + ss.ss_sp =0 ; + }else { + ret = -TARGET_ENOMEM ; +if (ss.ss_size < minstacksize) { + goto out; + } + } + + ts->sigaltstack_used.ss_sp = ss.ss_sp; + ts->sigaltstack_used.ss_size = ss.ss_size; + } + +if (uoss_addr) { + ret = -TARGET_EFAULT ; +if (copy_to_user (uoss_addr, &oss,sizeof (oss))) { + goto out; + } + } + + ret =0 ; +out: +return ret; +} +/* do_sigaction() return host values and errnos */ int do_sigaction (int sig,const struct target_sigaction *act,struct target_sigaction *oact)
Signature · 1 line
2.33.1
Re: [PATCH v2 01/40] bsd-user: Complete FreeBSD siginfo
In reply to Warner LoshQuoted 6 lines · Warner Losh
Fill in the missing FreeBSD siginfo fields, and add some comments.
Signed-off-by Warner Losh<[email protected]>--- bsd-user/freebsd/target_os_siginfo.h | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)
r~
Re: [PATCH v2 02/40] bsd-user: Create setup_sigframe_arch to setup sigframe context
In reply to Warner LoshQuoted 9 lines · Warner Losh
Define
setup_sigframe_archwhose job it is to setup the mcontext for the sigframe. Implement for x86 to just call mcontext.Signed-off-by Warner Losh<[email protected]>--- bsd-user/freebsd/target_os_signal.h | 3 +++ bsd-user/i386/signal.c | 13 +++++++++++++ bsd-user/x86_64/signal.c | 13 +++++++++++++ 3 files changed, 29 insertions(+)
r~
Re: [PATCH v2 03/40] bsd-user/arm/signal.c: Implement setup_sigframe_arch for arm
In reply to Warner LoshQuoted 11 lines · Warner Losh
Fix the broken context setting for arm. FreeBSD's
get_mcontextdoes not fill in the vfp info. It's filled in insigframe(). This corresponds to the newsetup_sigframe_archwhich fills in mcontext, then adjusts it to point to the vfp context in the sigframe and fills in that context as well. Add pointer to where this code is done.Signed-off-by Warner Losh <[email protected]>--- bsd-user/arm/signal.c | 50 +++++++++++++++++++---------- bsd-user/freebsd/target_os_signal.h | 2 +- 2 files changed, 34 insertions(+), 18 deletions(-)
bsd-user/freebsd/target_os_signal.h line 10
Quoted 13 lines
bsd-user/freebsd/target_os_signal.h +1
-1@@ -4,7 +4,7 @@#include "target_os_siginfo.h" #include "target_arch_signal.h" -abi_long setup_sigframe_arch (CPUArchState *regs,abi_ulong frame_addr, +abi_long setup_sigframe_arch (CPUArchState *env,abi_ulong frame_addr,struct target_sigframe *frame,int flags);/* Compare to sys/signal.h */
Should be in patch 2.
r~
Re: [PATCH v2 04/40] bsd-user/arm/signal.c: get_mcontext should zero vfp data
In reply to Warner LoshQuoted 15 lines · Warner Losh
FreeBSD's
get_mcontextdoesn't return any vfp data. Instead, it zeros out the vfp feilds (and all the spare fields). Impelement this behavior. We're still missing thesysarch(ARM_GET_VFPCONTEXT)syscall, though.Signed-off-by Warner Losh <[email protected]>--- bsd-user/arm/signal.c | 8 ++++++++ 1 file changed, 8 insertions(+)bsd-user/arm/signal.c +0
-0
bsd-user/arm/signal.c abi_long get_mcontext(CPUARMState *env, target_mcontext_t *mcp, int flags) · line 115
Quoted 7 lines
bsd-user/arm/signal.c +4
-0@@ -109,6 +109,14 @@ abi_long get_mcontext(CPUARMState *env, target_mcontext_t *mcp, int flags) gr[TARGET_REG_LR ] =tswap32 (env->regs[14 ]); gr[TARGET_REG_PC ] =tswap32 (env->regs[15 ]); +/* +* FreeBSD's set_mcontext doesn't save VFP info, but blanks it out instead. +* Instead, sysarch(ARM_GET_VFPSTATE) is used instead. +*/
Could be rewritten with fewer "instead". You wanted get_mcontext. Otherwise,
r~
Re: [PATCH v2 05/40] bsd-user: Remove vestiges of signal queueing code
In reply to Warner LoshQuoted 10 lines · Warner Losh
bsd-user was copied from linux-user at a time when it queued signals. Remove those vestiges of thse code. Retain the init function, even though it's now empty since other stuff will likely be added there. Make it static since it's not called from outside of
main.cSigned-off-by Warner Losh<[email protected]>--- bsd-user/main.c | 9 +-------- bsd-user/qemu.h | 13 +------------ 2 files changed, 2 insertions(+), 20 deletions(-)
r~
Re: [PATCH v2 06/40] bsd-user: Bring in docs from linux-user for signal_pending
In reply to Warner LoshQuoted 6 lines · Warner Losh
This is currently unused, so no code adjustments are needed.
Signed-off-by Warner Losh<[email protected]>--- bsd-user/qemu.h | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
r~
Re: [PATCH v2 09/40] bsd-user/signal-common.h: Move signal functions prototypes to here
In reply to Warner LoshQuoted 5 lines · Warner Losh
Signed-off-by Warner Losh<[email protected]>--- bsd-user/qemu.h | 8 -------- bsd-user/signal-common.h | 6 ++++++ 2 files changed, 6 insertions(+), 8 deletions(-)
r~
Re: [PATCH v2 14/40] bsd-user/arm/target_arch_cpu.h: Use force_sig_fault for EXCP_UDEF
In reply to Warner Loshbsd-user/arm/target_arch_cpu.h static inline void target_cpu_loop(CPUARMState *env) · line 60
Quoted 7 lines · Warner Losh
bsd-user/arm/target_arch_cpu.h +7
-0+case EXCP_NOCP : +case EXCP_INVSTATE : +/* +* See arm/arm/undefined.c undefinedinstruction(); +* +* A number of details aren't emulated (they likely don't matter): +* o Misaligned PC generates ILL_ILLADR
As I mentioned, misaligned pc will not come here for qemu. In the Arm ARM, see aarch32/functions/registers/BXWritePC:
// For branches to an unaligned PC counter in A32 state, the processor takes the branch // and does one of:
// * Forces the address to be aligned
// * Leaves the PC unaligned, meaning the target generates a PC Alignment fault.
The hardware will either refuse to allow bit 1 to be set when bit 0 is clear, OR it will generate a PREFETCH_DATA_ABORT for Alignment.
QEMU will do the latter.
Otherwise,
r~
Re: [PATCH v2 16/40] bsd-user/signal.c: implement abstract target / host signal translation
In reply to Warner LoshQuoted 10 lines · Warner Losh
Implement
host_to_target_signalandtarget_to_host_signal.Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>Reviewed-by Peter Maydell<[email protected]>--- bsd-user/signal-common.h | 2 ++ bsd-user/signal.c | 16 ++++++++++++++++ 2 files changed, 18 insertions(+)
r~
Re: [PATCH v2 18/40] bsd-user/signal.c: Add si_type argument to queue_signal
In reply to Warner LoshQuoted 11 lines · Warner Losh
Mirror the linux-user practice and add a
si_typeargument to queue signal. This will be transported as the upper 8 bits in thesi_typeelement of siginfo so that we know what bits of the structure are valid and so we can properly implementhost_to_target_siginfo_noswapandtswap_siginfo. Adapt the one caller ofqueue_signalto the new interface. Use all the same names as Linux (except_RTwhich we don't treat differently, unlike Linux), though some are unused. Place this intosignal-common.hsince that's a better place given bsd-user's structure. Move prototype ofqueue_signaltosignal-common.hto mirror linux-user's location.
Quoted 5 lines
Signed-off-by Warner Losh<[email protected]>--- bsd-user/signal-common.h | 26 +++++++++++++++++++++++++- bsd-user/signal.c | 5 +++-- 2 files changed, 28 insertions(+), 3 deletions(-)
r~
Re: [PATCH v2 19/40] bsd-user/host/arm/host-signal.h: Implement host_signal_*
In reply to Warner LoshQuoted 9 lines · Warner Losh
Implement
host_signal_pc,host_signal_set_pcandhost_signal_writefor arm.Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/host/arm/host-signal.h | 35 +++++++++++++++++++++++++++++++++ 1 file changed, 35 insertions(+)create mode 100644
bsd-user/host/arm/host-signal.h
r~
Re: [PATCH v2 26/40] bsd-user/signal.c: Implement host_signal_handler
In reply to Warner LoshQuoted 9 lines · Warner Losh
Implement
host_signal_handlerto handle signals generated by the host and to do safe system calls.Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/signal.c | 105 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 105 insertions(+)
r~
Re: [PATCH v2 31/40] bsd-user/signal.c: setup_frame
In reply to Warner LoshQuoted 11 lines · Warner Losh
setup_framesets up a signalled stack frame. Associated routines to extract the pointer to the stack frame and to support alternate stacks.Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/main.c | 5 +++ bsd-user/qemu.h | 3 +- bsd-user/signal.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 89 insertions(+), 1 deletion(-)
bsd-user/signal.c static void host_signal_handler(int host_sig, siginfo_t *info, void *puc) · line 523
Quoted 7 lines
bsd-user/signal.c +7
-0+#if defined(TARGET_ARM ) +return (sp - frame_size) & ~7 ; +#elif defined(TARGET_AARCH64 ) +return (sp - frame_size) & ~15 ; +#else +return sp - frame_size; +#endif
Just double-checking that this is still in the cleanup queue. I would expect x86 to require 16 byte alignment as well, for sse.
r~
Re: [PATCH v2 29/40] bsd-user/signal.c: Fill in queue_signal
In reply to Warner LoshQuoted 10 lines · Warner Losh
Fill in queue signal implementation, as well as routines allocate and delete elements of the signal queue.
Signed-off-by Stacey Son <[email protected]>Signed-off-by Kyle Evans <[email protected]>Signed-off-by Warner Losh <[email protected]>--- bsd-user/qemu.h | 1 + bsd-user/signal.c | 13 ++++++++++++- 2 files changed, 13 insertions(+), 1 deletion(-)
Quoted 2 lines
+ struct emulated_sigtable sync_signal;
struct emulated_sigtable sigtab[TARGET_NSIG];I'll note that we don't need an array of these, since we block all signals while returning to the main cpu loop, so we can't receive a second async signal. Something to be fixed for both l-user and b-user later...
r~
Re: [PATCH v2 30/40] bsd-user/signal.c: sigset manipulation routines.
In reply to Warner LoshQuoted 8 lines · Warner Losh
target_sigemptyset: resets a set to having no bits settarget_sigaddset: adds a signal to a settarget_sigismember: returns true when signal is a memberhost_to_target_sigset_internal: convert host sigset to targethost_to_target_sigset: convert host sigset to targettarget_to_host_sigset_internal: convert target sigset to hosttarget_to_host_sigset: convert target sigset to host
Quoted 7 lines
Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/signal-common.h | 2 ++ bsd-user/signal.c | 74 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 76 insertions(+)
r~
Re: [PATCH v2 32/40] bsd-user/signal.c: handle_pending_signal
In reply to Warner LoshQuoted 9 lines · Warner Losh
Handle a queued signal.
Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/qemu.h | 7 ++++ bsd-user/signal.c | 87 +++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 94 insertions(+)
r~
Re: [PATCH v2 34/40] bsd-user/signal.c: process_pending_signals
In reply to Warner LoshQuoted 8 lines · Warner Losh
Process the currently queued signals.
Signed-off-by Stacey Son <[email protected]>Signed-off-by Kyle Evans <[email protected]>Signed-off-by Warner Losh <[email protected]>--- bsd-user/signal.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 55 insertions(+), 1 deletion(-)
Quoted 1 line
+ * us to take anothe rhost signal, which will set
signal_pendingagain.
Typo around "another".
r~
Re: [PATCH v2 40/40] bsd-user/freebsd/target_os_ucontext.h: Prefer env as arg name for CPUArchState args
In reply to Warner LoshQuoted 4 lines · Warner Losh
Signed-off-by Warner Losh<[email protected]>--- bsd-user/freebsd/target_os_ucontext.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
r~
Re: [PATCH v2 24/40] bsd-user/signal.c: host_to_target_siginfo_noswap
In reply to Warner LoshQuoted 8 lines · Warner Losh
Implement conversion of host to target siginfo.
Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/signal.c | 113 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 113 insertions(+)
r~
Re: [PATCH v2 39/40] bsd-user: Rename arg name for target_cpu_reset to env
In reply to Warner LoshQuoted 9 lines · Warner Losh
Rename the parameter name for target_cpu_reset's CPUArchState * arg from cpu to env.
Signed-off-by Warner Losh<[email protected]>--- bsd-user/arm/target_arch_cpu.h | 2 +- bsd-user/i386/target_arch_cpu.h | 4 ++-- bsd-user/x86_64/target_arch_cpu.h | 4 ++-- 3 files changed, 5 insertions(+), 5 deletions(-)
r~
Re: [PATCH v2 37/40] bsd-user/signal.c: do_sigaltstack
In reply to Warner LoshQuoted 11 lines · Warner Losh
Implement the meat of the sigaltstack(2) system call with
do_sigaltstack.With that, all the stubbed out routines are complete, so remove now-incorrect comment.
Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/signal.c | 72 +++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 67 insertions(+), 5 deletions(-)
r~
Re: [PATCH v2 33/40] bsd-user/signal.c: tswap_siginfo
In reply to Warner LoshQuoted 8 lines · Warner Losh
Convert siginfo from targer to host.
Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/signal.c | 53 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 53 insertions(+)
r~
Re: [PATCH v2 38/40] MAINTAINERS: Add tests/vm/*bsd to the list to get reviews on
In reply to Warner LoshQuoted 7 lines · Warner Losh
tests/vm/*bsd (especailly tests/vm/freebsd) are adjacent to the bsd-user stuff and we're keen on keeping them working as well.
Signed-off-by Warner Losh<[email protected]>--- MAINTAINERS | 1 + 1 file changed, 1 insertion(+)
r~
Re: [PATCH v2 14/40] bsd-user/arm/target_arch_cpu.h: Use force_sig_fault for EXCP_UDEF
In reply to Richard HendersonQuoted 20 lines · Richard Henderson
bsd-user/arm/target_arch_cpu.h static inline void target_cpu_loop(CPUARMState *env)
Quoted 6 lines · Warner Losh
bsd-user/arm/target_arch_cpu.h +6
-0+case EXCP_NOCP : +case EXCP_INVSTATE : +/* +* See arm/arm/undefined.c undefinedinstruction(); +* +* A number of details aren't emulated (they likely don't matter):
Quoted 1 line
+ * o Misaligned PC generates
ILL_ILLADRAs I mentioned, misaligned pc will not come here for qemu. In the Arm ARM, see aarch32/functions/registers/BXWritePC:
// For branches to an unaligned PC counter in A32 state, the processor takes the branch
// and does one of:
// * Forces the address to be aligned
// * Leaves the PC unaligned, meaning the target generates a PC Alignment fault.
Quoted 6 lines
The hardware will either refuse to allow bit 1 to be set when bit 0 is clear, OR it will
generate aPREFETCH_DATA_ABORTfor Alignment.QEMU will do the latter.
Ah, right. I'd meant to update the comments and it slipped my mind. I'll note
that this can't happen in qemu.
Quoted 4 lines
Otherwise,
Reviewed-by Richard Henderson <[email protected]>
Thanks for this (and all the other) review and feedback.
Warner
Quoted 2 lines
r~
Re: [PATCH v2 31/40] bsd-user/signal.c: setup_frame
In reply to Richard HendersonQuoted 16 lines · Richard Henderson
Quoted 11 lines · Warner Losh
setup_framesets up a signalled stack frame. Associated routines to extract the pointer to the stack frame and to support alternate stacks.Signed-off-by Stacey Son<[email protected]>Signed-off-by Kyle Evans<[email protected]>Signed-off-by Warner Losh<[email protected]>--- bsd-user/main.c | 5 +++ bsd-user/qemu.h | 3 +- bsd-user/signal.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 89 insertions(+), 1 deletion(-)Reviewed-by Richard Henderson <[email protected]>
Quoted 10 lines
bsd-user/signal.c static void host_signal_handler(int host_sig, siginfo_t *info, void *puc) · line 523
Quoted 7 lines
bsd-user/signal.c +7
-0+#if defined(TARGET_ARM ) +return (sp - frame_size) & ~7 ; +#elif defined(TARGET_AARCH64 ) +return (sp - frame_size) & ~15 ; +#else +return sp - frame_size; +#endif Just double-checking that this is still in the cleanup queue. I would expect x86 to require 16 byte alignment as well, for sse
I'll add
/* TODO: make this a target_arch function / define */
to make it clear that it's still on the list... Thanks for the reminder. There's
a lot on the list, alas...
Warner
Quoted 3 lines
r~
Re: [PATCH v2 04/40] bsd-user/arm/signal.c: get_mcontext should zero vfp data
In reply to Richard HendersonQuoted 20 lines · Richard Henderson
bsd-user/arm/signal.c abi_long get_mcontext(CPUARMState *env,
Quoted 15 lines · Warner Losh
FreeBSD's
get_mcontextdoesn't return any vfp data. Instead, it zeros out the vfp feilds (and all the spare fields). Impelement this behavior. We're still missing thesysarch(ARM_GET_VFPCONTEXT)syscall, though.Signed-off-by Warner Losh <[email protected]>--- bsd-user/arm/signal.c | 8 ++++++++ 1 file changed, 8 insertions(+)bsd-user/arm/signal.c +0
-0@@ -109,6 +109,14 @@ abi_long get_mcontext(CPUARMState *env,
target_mcontext_t*mcp, int flags)Quoted 3 lines
bsd-user/arm/signal.c +0
-0gr[TARGET_REG_LR ] =tswap32 (env->regs[14 ]); gr[TARGET_REG_PC ] =tswap32 (env->regs[15 ]);
Quoted 8 lines
bsd-user/arm/signal.c abi_long get_mcontext(CPUARMState *env, · line 113
Quoted 2 lines
bsd-user/arm/signal.c +2
-0+/* +* FreeBSD's set_mcontext doesn't save VFP info, but blanks it out instead.
bsd-user/arm/signal.c abi_long get_mcontext(CPUARMState *env, target_mcontext_t *mcp, int flags) · line 115
Quoted 2 lines
bsd-user/arm/signal.c +2
-0+* Instead, sysarch(ARM_GET_VFPSTATE) is used instead. +*/ Could be rewritten with fewer "instead". You wanted
get_mcontext.
Yes. Instead is indeed over used over and over redundantly :)
Thanks!
Warner
Quoted 8 lines
Otherwise,
Reviewed-by Richard Henderson <[email protected]>r~
Re: [PATCH v2 29/40] bsd-user/signal.c: Fill in queue_signal
In reply to Richard HendersonQuoted 14 lines · Richard Henderson
Quoted 10 lines · Warner Losh
Fill in queue signal implementation, as well as routines allocate and delete elements of the signal queue.
Signed-off-by Stacey Son <[email protected]>Signed-off-by Kyle Evans <[email protected]>Signed-off-by Warner Losh <[email protected]>--- bsd-user/qemu.h | 1 + bsd-user/signal.c | 13 ++++++++++++- 2 files changed, 13 insertions(+), 1 deletion(-)Reviewed-by Richard Henderson <[email protected]>
Quoted 9 lines
Quoted 2 lines
+ struct emulated_sigtable sync_signal; struct emulated_sigtable sigtab[TARGET_NSIG];I'll note that we don't need an array of these, since we block all signals while returning
to the main cpu loop, so we can't receive a second async signal. Something to be fixed
for both l-user and b-user later...
I'll add
+ /*
+ * TODO: Since we block all signals while returning to the main CPU + * loop, this needn't be an array + */
before the array to document this so it doesn't get lost...
Warner
Re: [PATCH v2 03/40] bsd-user/arm/signal.c: Implement setup_sigframe_arch for arm
In reply to Richard HendersonQuoted 24 lines · Richard Henderson
Quoted 11 lines · Warner Losh
Fix the broken context setting for arm. FreeBSD's
get_mcontextdoes not fill in the vfp info. It's filled in insigframe(). This corresponds to the newsetup_sigframe_archwhich fills in mcontext, then adjusts it to point to the vfp context in the sigframe and fills in that context as well. Add pointer to where this code is done.Signed-off-by Warner Losh <[email protected]>--- bsd-user/arm/signal.c | 50 +++++++++++++++++++---------- bsd-user/freebsd/target_os_signal.h | 2 +- 2 files changed, 34 insertions(+), 18 deletions(-)Reviewed-by Richard Henderson <[email protected]>Review
Quoted 1 line
(quoted diff) +0
-0
b/bsd-user/freebsd/target_os_signal.hQuoted 7 lines
index
7491629477a..43700d08f71100644bsd-user/freebsd/target_os_signal.h +0
-0@@ -4,7 +4,7 @@#include "target_os_siginfo.h" #include "target_arch_signal.h"
Quoted 8 lines
bsd-user/freebsd/target_os_signal.h line 10
Quoted 5 lines
bsd-user/freebsd/target_os_signal.h +1
-1-abi_long setup_sigframe_arch (CPUArchState *regs,abi_ulong frame_addr, +abi_long setup_sigframe_arch (CPUArchState *env,abi_ulong frame_addr,struct target_sigframe *frame,int flags);/* Compare to sys/signal.h */ Should be in patch 2.
Yea, I back merged those changes after making them on blitz and landed it in the
wrong spot. Good eye.
Warner
Quoted 2 lines
r~
Re: [PATCH v2 36/40] bsd-user/signal.c: implement do_sigaction
In reply to Warner LoshQuoted 14 lines · Warner Losh
Implement the meat of the sigaction(2) system call with
do_sigactionand helper routinerblock_signals(which is also used to implemement signal masking so it's global).Signed-off-by Stacey Son <[email protected]>Signed-off-by Kyle Evans <[email protected]>Signed-off-by Warner Losh <[email protected]>Pending Comments from Peter Maydell <[email protected]>
Quoted 8 lines
(1) in
block_signals, sigprocmast
For linux-user we rely onsigprocmask()in a multithreaded program setting the signal mask for only the calling thread, which isn't POSIX-mandated. (Arguably we should usepthread_sigmask()instead, but we don't for basically historical reasons since linux-user is host-OS-specific anyway.) Does BSD have the same "this changes this thread's signal mask" semantics forsigprocmask()?
FreeBSD changes this on a per-thread basis for both sigprocmask and pthread_sigmask(). pthread_sigmask() just does some extra stuff with SIGCANCEL for pthread_cancel support which qemu doesn’t use. They are the same. I’m inclined to leave it as sigprocmask() since I’m unsure what the implications of doing funky things for SIGCANCEL would be.
Quoted 5 lines
(2)
do_sigaction, first if validating which signals can be sent Kernel seems to allow SIGKILL and SIGSTOP unless act is non-NULL andact->sa_handlerisSIG_DFL?
https://github.com/freebsd/freebsd-src/blob/main/sys/kern/kern_sig.c#L747
(Compare linux-user commitee3500d33a7431, a recent bugfix.)
That fix is relevant, so I’ll bring that in. Thanks!
Quoted 3 lines
(3) Noting confusion in
do_sigactionbetween host and target errnos (they are identical on BSD, but we should still return the right sort in case it ever does matter).
Will fix before V3 of patches.
Warner
Quoted 136 lines
---
bsd-user/signal-common.h| 22 ++++++++++++bsd-user/signal.c| 76 ++++++++++++++++++++++++++++++++++++++++2 files changed, 98 insertions(+)bsd-user/signal-common.h +22
-0@@ -9,7 +9,29 @@#ifndef SIGNAL_COMMON_H #define SIGNAL_COMMON_H +/** +* block_signals: block all signals while handling this guest syscall +* +* Block all signals, and arrange that the signal mask is returned to +* its correct value for the guest before we resume execution of guest code. +* If this function returns non-zero, then the caller should immediately +* return -TARGET_ERESTARTSYS to the main loop, which will take the pending +* signal and restart execution of the syscall. +* If block_signals() returns zero, then the caller can continue with +* emulation of the system call knowing that no signals can be taken +* (and therefore that no race conditions will result). +* This should only be called once, because if it is called a second time +* it will always return non-zero. (Think of it like a mutex that can't +* be recursively locked.) +* Signals will be unblocked again by process_pending_signals(). +* +* Return value: non-zero if there was a pending signal, zero if not. +*/ +int block_signals (void );/* Returns non zero if signal pending */ +long do_rt_sigreturn (CPUArchState *env); +int do_sigaction (int sig,const struct target_sigaction *act, +struct target_sigaction *oact);abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp);long do_sigreturn (CPUArchState *env,abi_ulong addr);void force_sig_fault (int sig,int code,abi_ulong addr);bsd-user/signal.c +76
-0@@ -309,6 +309,22 @@ static void tswap_siginfo(target_siginfo_t *tinfo, const target_siginfo_t *info) } } +int block_signals (void ) +{ +TaskState *ts = (TaskState *)thread_cpu->opaque; +sigset_t set; + +/* +* It's OK to block everything including SIGSEGV, because we won't run any +* further guest code before unblocking signals in +* process_pending_signals(). +*/ +sigfillset (&set); +sigprocmask (SIG_SETMASK , &set,0 ); + +return qatomic_xchg (&ts->signal_pending,1 ); +} +/* Returns 1 if given signal should dump core if not handled. */ static int core_dump_signal (int sig) {@@ -554,6 +570,66 @@ static void host_signal_handler(int host_sig, siginfo_t *info, void *puc)cpu_exit (thread_cpu); } +/* do_sigaction() return host values and errnos */ +int do_sigaction (int sig,const struct target_sigaction *act, +struct target_sigaction *oact) +{ +struct target_sigaction *k; +struct sigaction act1; +int host_sig; +int ret =0 ; + +if (sig <1 || sig >TARGET_NSIG || sig ==TARGET_SIGKILL || + sig ==TARGET_SIGSTOP ) { +return -EINVAL ; + } + +if (block_signals ()) { +return -TARGET_ERESTART ; + } + + k = &sigact_table[sig -1 ]; +if (oact) { + oact->_sa_handler =tswapal (k->_sa_handler); + oact->sa_flags =tswap32 (k->sa_flags); + oact->sa_mask = k->sa_mask; + } +if (act) { +/* XXX: this is most likely not threadsafe. */ + k->_sa_handler =tswapal (act->_sa_handler); + k->sa_flags =tswap32 (act->sa_flags); + k->sa_mask = act->sa_mask; + +/* Update the host signal state. */ + host_sig =target_to_host_signal (sig); +if (host_sig !=SIGSEGV && host_sig !=SIGBUS ) { +memset (&act1,0 ,sizeof (struct sigaction )); +sigfillset (&act1.sa_mask); + act1.sa_flags =SA_SIGINFO ; +if (k->sa_flags &TARGET_SA_RESTART ) { + act1.sa_flags |=SA_RESTART ; + } +/* +* Note: It is important to update the host kernel signal mask to +* avoid getting unexpected interrupted system calls. +*/ +if (k->_sa_handler ==TARGET_SIG_IGN ) { + act1.sa_sigaction = (void *)SIG_IGN ; + }else if (k->_sa_handler ==TARGET_SIG_DFL ) { +if (fatal_signal (sig)) { + act1.sa_sigaction = host_signal_handler; + }else { + act1.sa_sigaction = (void *)SIG_DFL ; + } + }else { + act1.sa_sigaction = host_signal_handler; + } + ret =sigaction (host_sig, &act1,NULL ); + } + } +return ret; +} +static inline abi_ulong get_sigframe (struct target_sigaction *ka,CPUArchState *env,size_t frame_size) {Signature · 1 line
2.33.1
Re: [PATCH v2 35/40] bsd-user/signal.c: implement do_sigreturn
In reply to Warner LoshQuoted 14 lines · Warner Losh
Implements the meat of a sigreturn(2) system call via
do_sigreturn, and helperreset_signal_mask. Fix the prototype ofdo_sigreturninqemu.hand removedo_rt_sigreturnsince it's linux only.Signed-off-by Stacey Son <[email protected]>Signed-off-by Kyle Evans <[email protected]>Signed-off-by Warner Losh <[email protected]>PENDING COMMENTS: from Peter Maydell <[email protected]>
Quoted 9 lines
do_sigreturn()itself shouldn't be setting the active signal mask, at least if you follow the linux-user design. It just sets the thread'ssignal_maskfield in the TaskState by callingset_sigmask(), and then on our way out in the main cpu loop we'll callprocess_pending_signals()which sets the real thread signal mask to that value. (This, together withdo_sigreturn()callingblock_signals()before it starts work, avoids some race conditions where a host signal is delivered as soon as we unblock, I think.)
I think this is correct. We’re already setting the signal_mask element in a mostly similar way to linux-user, so all I need to do is to remove the setting the actual signal mask. The rest of this is already in place. I’ll included in v3.
Warner
Quoted 88 lines
---
bsd-user/signal-common.h| 2 +-bsd-user/signal.c| 56 ++++++++++++++++++++++++++++++++++++++++2 files changed, 57 insertions(+), 1 deletion(-)bsd-user/signal-common.h +1
-1@@ -11,7 +11,7 @@long do_rt_sigreturn (CPUArchState *env);abi_long do_sigaltstack (abi_ulong uss_addr,abi_ulong uoss_addr,abi_ulong sp); -long do_sigreturn (CPUArchState *env); +long do_sigreturn (CPUArchState *env,abi_ulong addr);void force_sig_fault (int sig,int code,abi_ulong addr);int host_to_target_signal (int sig);void host_to_target_sigset (target_sigset_t *d,const sigset_t *s);bsd-user/signal.c +56
-0@@ -626,6 +626,62 @@ static void setup_frame(int sig, int code, struct target_sigaction *ka,unlock_user_struct (frame, frame_addr,1 ); } +static int reset_signal_mask (target_ucontext_t *ucontext) +{ +int i; +sigset_t blocked; +target_sigset_t target_set; +TaskState *ts = (TaskState *)thread_cpu->opaque; + +for (i =0 ; i <TARGET_NSIG_WORDS ; i++) +if (__get_user (target_set.__bits[i], + &ucontext->uc_sigmask.__bits[i])) { +return -TARGET_EFAULT ; + } +target_to_host_sigset_internal (&blocked, &target_set); + ts->signal_mask = blocked; +sigprocmask (SIG_SETMASK , &ts->signal_mask,NULL ); + +return 0 ; +} + +/* See sys/$M/$M/exec_machdep.c sigreturn() */ +long do_sigreturn (CPUArchState *env,abi_ulong addr) +{ +long ret; +abi_ulong target_ucontext; +target_ucontext_t *ucontext =NULL ; + +/* Get the target ucontext address from the stack frame */ + ret =get_ucontext_sigreturn (env, addr, &target_ucontext); +if (is_error (ret)) { +return ret; + } +trace_user_do_sigreturn (env, addr); +if (!lock_user_struct (VERIFY_READ , ucontext, target_ucontext,0 )) { + goto badframe; + } + +/* Set the register state back to before the signal. */ +if (set_mcontext (env, &ucontext->uc_mcontext,1 )) { + goto badframe; + } + +/* And reset the signal mask. */ +if (reset_signal_mask (ucontext)) { + goto badframe; + } + +unlock_user_struct (ucontext, target_ucontext,0 ); +return -TARGET_EJUSTRETURN ; + +badframe: +if (ucontext !=NULL ) { +unlock_user_struct (ucontext, target_ucontext,0 ); + } +return -TARGET_EFAULT ; +} +void signal_init (void ) {TaskState *ts = (TaskState *)thread_cpu->opaque;Signature · 1 line
2.33.1
Re: [PATCH v2 36/40] bsd-user/signal.c: implement do_sigaction
In reply to Warner LoshQuoted 24 lines · Warner Losh
Quoted 15 lines · Warner Losh
Implement the meat of the sigaction(2) system call with
do_sigactionand helper routinerblock_signals(which is also used to implemement signal masking so it's global).Signed-off-by Stacey Son <[email protected]>Signed-off-by Kyle Evans <[email protected]>Signed-off-by Warner Losh <[email protected]>Pending Comments from Peter Maydell <[email protected]>
Quoted 8 lines
(1) in
block_signals, sigprocmast
For linux-user we rely onsigprocmask()in a multithreaded program setting the signal mask for only the calling thread, which isn't POSIX-mandated. (Arguably we should usepthread_sigmask()instead, but we don't for basically historical reasons since linux-user is host-OS-specific anyway.) Does BSD have the same "this changes this thread's signal mask" semantics forsigprocmask()?
Quoted 6 lines
FreeBSD changes this on a per-thread basis for both sigprocmask and
pthread_sigmask().pthread_sigmask()just does some extra stuff with SIGCANCEL forpthread_cancelsupport which qemu doesn’t use. They are the same. I’m inclined to leave it assigprocmask()since I’m unsure what the implications of doing funky things for SIGCANCEL would be.
Yes, that sounds like the right thing.
thanks
-- PMM