[PATCH v4 0/2] linux-user: fixes for sched_ syscalls

[PATCH v4 0/2] linux-user: fixes for sched_ syscalls

[Tonis Tiigi]

This patchset improves support for sched_* syscalls under user emulation.

The first commit adds support for sched_g/setattr that was previously not 
implemented. These syscalls are not exposed by glibc. The struct type needs
to be redefined as it can't be included directly before
https://lkml.org/lkml/2020/5/28/810 .

sched_attr type can grow in future kernel versions. When client sends
values that QEMU does not understand it will return E2BIG with same
semantics as old kernel would so client can retry with smaller inputs.

The second commit fixes sched_g/setscheduler and sched_g/setparam, when QEMU is 
built with musl. Musl does not implement these due to conflict between what 
these functions should do in syscalls and libc 
https://git.musl-libc.org/cgit/musl/commit/?id=1e21e78bf7a5c24c217446d8760be7b7188711c2
 . I've changed it to call syscall directly what should always be the expected 
behavior for the user.

Via https://github.com/tonistiigi/binfmt/pull/70
https://github.com/tonistiigi/binfmt/pull/73 with additional tests.


Changes v4->v3:
- host `sched_param` type is used for local syscalls
- Added check_zeroed_user() helper. This function takes kernel and userspace size
and checks if the extra padding is empty with same rules as kernel does. I also
tried with only one size parameter but doing all the `size-sizeof()` calculations
on caller side made the function quite useless.
- Defined local host version for `sched_param` so target and host types are defined
separately.
- Moved size type declaration to the beginning of the function.

Changes v3->v2:
- Fix wrong property name for sched_flags
- Validate size parameter and handle E2BIG errors same way as kernel does. There
is one case where it can't be done completely correctly but clients should still
be able to handle it: when client sends a bigger non-zero structure than current
kernel definition we will send E2BIG with the struct size known to qemu. If now
the client sends structure with this size it may still get another E2BIG error
from the kernel if kernel is old and doesn't implement util_min/util_max. I don't
think this can be handled without making additional syscalls to kernel.

Changes v1->v2:
- Locking guest addresses for sched_attr is now based on size inputs, not local 
struct size. Also did the same for setter where I now read only the size field 
of the struct first.
- Use offsetof() when checking if optional fields are supported.
- `target_sched_attr` now uses aligned types as requested. I didn't quite 
understand why this is needed as I don't see same in kernel headers, but as 
this type uses only constant width fields and is already aligned by default it 
can't break anything.
- Fixed formatting.
- Defined own `target_sched_param` struct as requested.

Tonis Tiigi (2):
  linux-user: add sched_getattr support
  linux-user: call set/getscheduler set/getparam directly

 linux-user/syscall.c      | 157 +++++++++++++++++++++++++++++++++++---
 linux-user/syscall_defs.h |  18 +++++
 2 files changed, 165 insertions(+), 10 deletions(-)

-- 
2.32.0 (Apple Git-132)

[PATCH v4 1/2] linux-user: add sched_getattr support

[Tonis Tiigi]

These syscalls are not exposed by glibc. The struct type need to be
redefined as it can't be included directly before
https://lkml.org/lkml/2020/5/28/810 .

sched_attr type can grow in future kernel versions. When client sends
values that QEMU does not understand it will return E2BIG with same
semantics as old kernel would so client can retry with smaller inputs.

Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
---
 linux-user/syscall.c      | 123 ++++++++++++++++++++++++++++++++++++++
 linux-user/syscall_defs.h |  14 +++++
 2 files changed, 137 insertions(+)

diff --git a/linux-user/syscall.c b/linux-user/syscall.c
index 56a3e17183..1b8415c8a3 100644
--- a/linux-user/syscall.c
+++ b/linux-user/syscall.c
@@ -340,6 +340,25 @@ _syscall3(int, sys_sched_getaffinity, pid_t, pid, unsigned int, len,
 #define __NR_sys_sched_setaffinity __NR_sched_setaffinity
 _syscall3(int, sys_sched_setaffinity, pid_t, pid, unsigned int, len,
           unsigned long *, user_mask_ptr);
+/* sched_attr is not defined in glibc */
+struct sched_attr {
+    uint32_t size;
+    uint32_t sched_policy;
+    uint64_t sched_flags;
+    int32_t sched_nice;
+    uint32_t sched_priority;
+    uint64_t sched_runtime;
+    uint64_t sched_deadline;
+    uint64_t sched_period;
+    uint32_t sched_util_min;
+    uint32_t sched_util_max;
+};
+#define __NR_sys_sched_getattr __NR_sched_getattr
+_syscall4(int, sys_sched_getattr, pid_t, pid, struct sched_attr *, attr,
+          unsigned int, size, unsigned int, flags);
+#define __NR_sys_sched_setattr __NR_sched_setattr
+_syscall3(int, sys_sched_setattr, pid_t, pid, struct sched_attr *, attr,
+          unsigned int, flags);
 #define __NR_sys_getcpu __NR_getcpu
 _syscall3(int, sys_getcpu, unsigned *, cpu, unsigned *, node, void *, tcache);
 _syscall4(int, reboot, int, magic1, int, magic2, unsigned int, cmd,
@@ -558,6 +577,24 @@ const char *target_strerror(int err)
     return strerror(target_to_host_errno(err));
 }
 
+static int check_zeroed_user(abi_long addr, size_t ksize, size_t usize)
+{
+    int i;
+    uint8_t b;
+    if (usize <= ksize) {
+        return 1;
+    }
+    for (i = ksize; i < usize; i++) {
+        if (get_user_u8(b, addr + i)) {
+            return -TARGET_EFAULT;
+        }
+        if (b != 0) {
+            return 0;
+        }
+    }
+    return 1;
+}
+
 #define safe_syscall0(type, name) \
 static type safe_##name(void) \
 { \
@@ -10594,6 +10631,92 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
         }
     case TARGET_NR_sched_getscheduler:
         return get_errno(sched_getscheduler(arg1));
+    case TARGET_NR_sched_getattr:
+        {
+            struct target_sched_attr *target_scha;
+            struct sched_attr scha;
+            if (arg2 == 0) {
+                return -TARGET_EINVAL;
+            }
+            if (arg3 > sizeof(scha)) {
+                arg3 = sizeof(scha);
+            }
+            ret = get_errno(sys_sched_getattr(arg1, &scha, arg3, arg4));
+            if (!is_error(ret)) {
+                target_scha = lock_user(VERIFY_WRITE, arg2, arg3, 0);
+                if (!target_scha) {
+                    return -TARGET_EFAULT;
+                }
+                target_scha->size = tswap32(scha.size);
+                target_scha->sched_policy = tswap32(scha.sched_policy);
+                target_scha->sched_flags = tswap64(scha.sched_flags);
+                target_scha->sched_nice = tswap32(scha.sched_nice);
+                target_scha->sched_priority = tswap32(scha.sched_priority);
+                target_scha->sched_runtime = tswap64(scha.sched_runtime);
+                target_scha->sched_deadline = tswap64(scha.sched_deadline);
+                target_scha->sched_period = tswap64(scha.sched_period);
+                if (scha.size > offsetof(struct sched_attr, sched_util_min)) {
+                    target_scha->sched_util_min = tswap32(scha.sched_util_min);
+                    target_scha->sched_util_max = tswap32(scha.sched_util_max);
+                }
+                unlock_user(target_scha, arg2, arg3);
+            }
+            return ret;
+        }
+    case TARGET_NR_sched_setattr:
+        {
+            struct target_sched_attr *target_scha;
+            struct sched_attr scha;
+            uint32_t size;
+            int zeroed;
+            if (arg2 == 0) {
+                return -TARGET_EINVAL;
+            }
+            if (get_user_u32(size, arg2)) {
+                return -TARGET_EFAULT;
+            }
+            if (!size) {
+                size = offsetof(struct target_sched_attr, sched_util_min);
+            }
+            if (size < offsetof(struct target_sched_attr, sched_util_min)) {
+                if (put_user_u32(sizeof(struct target_sched_attr), arg2)) {
+                    return -TARGET_EFAULT;
+                }
+                return -TARGET_E2BIG;
+            }
+
+            zeroed = check_zeroed_user(arg2, sizeof(struct target_sched_attr), size);
+            if (zeroed < 0) {
+                return zeroed;
+            } else if (zeroed == 0) {
+                if (put_user_u32(sizeof(struct target_sched_attr), arg2)) {
+                    return -TARGET_EFAULT;
+                }
+                return -TARGET_E2BIG;
+            }
+            if (size > sizeof(struct target_sched_attr)) {
+                size = sizeof(struct target_sched_attr);
+            }
+
+            target_scha = lock_user(VERIFY_READ, arg2, size, 1);
+            if (!target_scha) {
+                return -TARGET_EFAULT;
+            }
+            scha.size = size;
+            scha.sched_policy = tswap32(target_scha->sched_policy);
+            scha.sched_flags = tswap64(target_scha->sched_flags);
+            scha.sched_nice = tswap32(target_scha->sched_nice);
+            scha.sched_priority = tswap32(target_scha->sched_priority);
+            scha.sched_runtime = tswap64(target_scha->sched_runtime);
+            scha.sched_deadline = tswap64(target_scha->sched_deadline);
+            scha.sched_period = tswap64(target_scha->sched_period);
+            if (size > offsetof(struct target_sched_attr, sched_util_min)) {
+                scha.sched_util_min = tswap32(target_scha->sched_util_min);
+                scha.sched_util_max = tswap32(target_scha->sched_util_max);
+            }
+            unlock_user(target_scha, arg2, 0);
+            return get_errno(sys_sched_setattr(arg1, &scha, arg3));
+        }
     case TARGET_NR_sched_yield:
         return get_errno(sched_yield());
     case TARGET_NR_sched_get_priority_max:
diff --git a/linux-user/syscall_defs.h b/linux-user/syscall_defs.h
index 0b13975937..310d6ce8ad 100644
--- a/linux-user/syscall_defs.h
+++ b/linux-user/syscall_defs.h
@@ -2914,4 +2914,18 @@ struct target_statx {
    /* 0x100 */
 };
 
+/* from kernel's include/linux/sched/types.h */
+struct target_sched_attr {
+    abi_uint size;
+    abi_uint sched_policy;
+    abi_ullong sched_flags;
+    abi_int sched_nice;
+    abi_uint sched_priority;
+    abi_ullong sched_runtime;
+    abi_ullong sched_deadline;
+    abi_ullong sched_period;
+    abi_uint sched_util_min;
+    abi_uint sched_util_max;
+};
+
 #endif
-- 
2.32.0 (Apple Git-132)

[PATCH v4 2/2] linux-user: call set/getscheduler set/getparam directly

[Tonis Tiigi]

There seems to be difference in syscall and libc definition of these
methods and therefore musl does not implement them (1e21e78bf7). Call
syscall directly to ensure the behavior of the libc of user application,
not the libc that was used to build QEMU.

Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
---
 linux-user/syscall.c      | 34 ++++++++++++++++++++++++----------
 linux-user/syscall_defs.h |  4 ++++
 2 files changed, 28 insertions(+), 10 deletions(-)

diff --git a/linux-user/syscall.c b/linux-user/syscall.c
index 1b8415c8a3..30962155a6 100644
--- a/linux-user/syscall.c
+++ b/linux-user/syscall.c
@@ -359,6 +359,17 @@ _syscall4(int, sys_sched_getattr, pid_t, pid, struct sched_attr *, attr,
 #define __NR_sys_sched_setattr __NR_sched_setattr
 _syscall3(int, sys_sched_setattr, pid_t, pid, struct sched_attr *, attr,
           unsigned int, flags);
+#define __NR_sys_sched_getscheduler __NR_sched_getscheduler
+_syscall1(int, sys_sched_getscheduler, pid_t, pid);
+#define __NR_sys_sched_setscheduler __NR_sched_setscheduler
+_syscall3(int, sys_sched_setscheduler, pid_t, pid, int, policy,
+          const struct sched_param *, param);
+#define __NR_sys_sched_getparam __NR_sched_getparam
+_syscall2(int, sys_sched_getparam, pid_t, pid,
+          struct sched_param *, param);
+#define __NR_sys_sched_setparam __NR_sched_setparam
+_syscall2(int, sys_sched_setparam, pid_t, pid,
+          const struct sched_param *, param);
 #define __NR_sys_getcpu __NR_getcpu
 _syscall3(int, sys_getcpu, unsigned *, cpu, unsigned *, node, void *, tcache);
 _syscall4(int, reboot, int, magic1, int, magic2, unsigned int, cmd,
@@ -10587,30 +10598,32 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
         return ret;
     case TARGET_NR_sched_setparam:
         {
-            struct sched_param *target_schp;
+            struct target_sched_param *target_schp;
             struct sched_param schp;
 
             if (arg2 == 0) {
                 return -TARGET_EINVAL;
             }
-            if (!lock_user_struct(VERIFY_READ, target_schp, arg2, 1))
+            if (!lock_user_struct(VERIFY_READ, target_schp, arg2, 1)) {
                 return -TARGET_EFAULT;
+            }
             schp.sched_priority = tswap32(target_schp->sched_priority);
             unlock_user_struct(target_schp, arg2, 0);
-            return get_errno(sched_setparam(arg1, &schp));
+            return get_errno(sys_sched_setparam(arg1, &schp));
         }
     case TARGET_NR_sched_getparam:
         {
-            struct sched_param *target_schp;
+            struct target_sched_param *target_schp;
             struct sched_param schp;
 
             if (arg2 == 0) {
                 return -TARGET_EINVAL;
             }
-            ret = get_errno(sched_getparam(arg1, &schp));
+            ret = get_errno(sys_sched_getparam(arg1, &schp));
             if (!is_error(ret)) {
-                if (!lock_user_struct(VERIFY_WRITE, target_schp, arg2, 0))
+                if (!lock_user_struct(VERIFY_WRITE, target_schp, arg2, 0)) {
                     return -TARGET_EFAULT;
+                }
                 target_schp->sched_priority = tswap32(schp.sched_priority);
                 unlock_user_struct(target_schp, arg2, 1);
             }
@@ -10618,19 +10631,20 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
         return ret;
     case TARGET_NR_sched_setscheduler:
         {
-            struct sched_param *target_schp;
+            struct target_sched_param *target_schp;
             struct sched_param schp;
             if (arg3 == 0) {
                 return -TARGET_EINVAL;
             }
-            if (!lock_user_struct(VERIFY_READ, target_schp, arg3, 1))
+            if (!lock_user_struct(VERIFY_READ, target_schp, arg3, 1)) {
                 return -TARGET_EFAULT;
+            }
             schp.sched_priority = tswap32(target_schp->sched_priority);
             unlock_user_struct(target_schp, arg3, 0);
-            return get_errno(sched_setscheduler(arg1, arg2, &schp));
+            return get_errno(sys_sched_setscheduler(arg1, arg2, &schp));
         }
     case TARGET_NR_sched_getscheduler:
-        return get_errno(sched_getscheduler(arg1));
+        return get_errno(sys_sched_getscheduler(arg1));
     case TARGET_NR_sched_getattr:
         {
             struct target_sched_attr *target_scha;
diff --git a/linux-user/syscall_defs.h b/linux-user/syscall_defs.h
index 310d6ce8ad..28b9fe9a47 100644
--- a/linux-user/syscall_defs.h
+++ b/linux-user/syscall_defs.h
@@ -2928,4 +2928,8 @@ struct target_sched_attr {
     abi_uint sched_util_max;
 };
 
+struct target_sched_param {
+    abi_int sched_priority;
+};
+
 #endif
-- 
2.32.0 (Apple Git-132)

Re: [PATCH v4 1/2] linux-user: add sched_getattr support

[Laurent Vivier]

Le 05/01/2022 à 05:18, Tonis Tiigi a écrit :
> These syscalls are not exposed by glibc. The struct type need to be
> redefined as it can't be included directly before
> https://lkml.org/lkml/2020/5/28/810 .
> 
> sched_attr type can grow in future kernel versions. When client sends
> values that QEMU does not understand it will return E2BIG with same
> semantics as old kernel would so client can retry with smaller inputs.
> 
> Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
> ---
>   linux-user/syscall.c      | 123 ++++++++++++++++++++++++++++++++++++++
>   linux-user/syscall_defs.h |  14 +++++
>   2 files changed, 137 insertions(+)
> 
> diff --git a/linux-user/syscall.c b/linux-user/syscall.c
> index 56a3e17183..1b8415c8a3 100644
> --- a/linux-user/syscall.c
> +++ b/linux-user/syscall.c
> @@ -340,6 +340,25 @@ _syscall3(int, sys_sched_getaffinity, pid_t, pid, unsigned int, len,
>   #define __NR_sys_sched_setaffinity __NR_sched_setaffinity
>   _syscall3(int, sys_sched_setaffinity, pid_t, pid, unsigned int, len,
>             unsigned long *, user_mask_ptr);
> +/* sched_attr is not defined in glibc */
> +struct sched_attr {
> +    uint32_t size;
> +    uint32_t sched_policy;
> +    uint64_t sched_flags;
> +    int32_t sched_nice;
> +    uint32_t sched_priority;
> +    uint64_t sched_runtime;
> +    uint64_t sched_deadline;
> +    uint64_t sched_period;
> +    uint32_t sched_util_min;
> +    uint32_t sched_util_max;
> +};
> +#define __NR_sys_sched_getattr __NR_sched_getattr
> +_syscall4(int, sys_sched_getattr, pid_t, pid, struct sched_attr *, attr,
> +          unsigned int, size, unsigned int, flags);
> +#define __NR_sys_sched_setattr __NR_sched_setattr
> +_syscall3(int, sys_sched_setattr, pid_t, pid, struct sched_attr *, attr,
> +          unsigned int, flags);
>   #define __NR_sys_getcpu __NR_getcpu
>   _syscall3(int, sys_getcpu, unsigned *, cpu, unsigned *, node, void *, tcache);
>   _syscall4(int, reboot, int, magic1, int, magic2, unsigned int, cmd,
> @@ -558,6 +577,24 @@ const char *target_strerror(int err)
>       return strerror(target_to_host_errno(err));
>   }
>   
> +static int check_zeroed_user(abi_long addr, size_t ksize, size_t usize)
> +{
> +    int i;
> +    uint8_t b;
> +    if (usize <= ksize) {
> +        return 1;
> +    }
> +    for (i = ksize; i < usize; i++) {
> +        if (get_user_u8(b, addr + i)) {
> +            return -TARGET_EFAULT;
> +        }
> +        if (b != 0) {
> +            return 0;
> +        }
> +    }
> +    return 1;
> +}
> +
>   #define safe_syscall0(type, name) \
>   static type safe_##name(void) \
>   { \
> @@ -10594,6 +10631,92 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
>           }
>       case TARGET_NR_sched_getscheduler:
>           return get_errno(sched_getscheduler(arg1));
> +    case TARGET_NR_sched_getattr:
> +        {
> +            struct target_sched_attr *target_scha;
> +            struct sched_attr scha;
> +            if (arg2 == 0) {
> +                return -TARGET_EINVAL;
> +            }
> +            if (arg3 > sizeof(scha)) {
> +                arg3 = sizeof(scha);
> +            }
> +            ret = get_errno(sys_sched_getattr(arg1, &scha, arg3, arg4));
> +            if (!is_error(ret)) {
> +                target_scha = lock_user(VERIFY_WRITE, arg2, arg3, 0);
> +                if (!target_scha) {
> +                    return -TARGET_EFAULT;
> +                }
> +                target_scha->size = tswap32(scha.size);
> +                target_scha->sched_policy = tswap32(scha.sched_policy);
> +                target_scha->sched_flags = tswap64(scha.sched_flags);
> +                target_scha->sched_nice = tswap32(scha.sched_nice);
> +                target_scha->sched_priority = tswap32(scha.sched_priority);
> +                target_scha->sched_runtime = tswap64(scha.sched_runtime);
> +                target_scha->sched_deadline = tswap64(scha.sched_deadline);
> +                target_scha->sched_period = tswap64(scha.sched_period);
> +                if (scha.size > offsetof(struct sched_attr, sched_util_min)) {
> +                    target_scha->sched_util_min = tswap32(scha.sched_util_min);
> +                    target_scha->sched_util_max = tswap32(scha.sched_util_max);
> +                }
> +                unlock_user(target_scha, arg2, arg3);
> +            }
> +            return ret;
> +        }
> +    case TARGET_NR_sched_setattr:
> +        {
> +            struct target_sched_attr *target_scha;
> +            struct sched_attr scha;
> +            uint32_t size;
> +            int zeroed;
> +            if (arg2 == 0) {
> +                return -TARGET_EINVAL;
> +            }
> +            if (get_user_u32(size, arg2)) {
> +                return -TARGET_EFAULT;
> +            }
> +            if (!size) {
> +                size = offsetof(struct target_sched_attr, sched_util_min);
> +            }
> +            if (size < offsetof(struct target_sched_attr, sched_util_min)) {
> +                if (put_user_u32(sizeof(struct target_sched_attr), arg2)) {
> +                    return -TARGET_EFAULT;
> +                }
> +                return -TARGET_E2BIG;
> +            }
> +
> +            zeroed = check_zeroed_user(arg2, sizeof(struct target_sched_attr), size);
> +            if (zeroed < 0) {
> +                return zeroed;
> +            } else if (zeroed == 0) {
> +                if (put_user_u32(sizeof(struct target_sched_attr), arg2)) {
> +                    return -TARGET_EFAULT;
> +                }
> +                return -TARGET_E2BIG;
> +            }
> +            if (size > sizeof(struct target_sched_attr)) {
> +                size = sizeof(struct target_sched_attr);
> +            }
> +
> +            target_scha = lock_user(VERIFY_READ, arg2, size, 1);
> +            if (!target_scha) {
> +                return -TARGET_EFAULT;
> +            }
> +            scha.size = size;
> +            scha.sched_policy = tswap32(target_scha->sched_policy);
> +            scha.sched_flags = tswap64(target_scha->sched_flags);
> +            scha.sched_nice = tswap32(target_scha->sched_nice);
> +            scha.sched_priority = tswap32(target_scha->sched_priority);
> +            scha.sched_runtime = tswap64(target_scha->sched_runtime);
> +            scha.sched_deadline = tswap64(target_scha->sched_deadline);
> +            scha.sched_period = tswap64(target_scha->sched_period);
> +            if (size > offsetof(struct target_sched_attr, sched_util_min)) {
> +                scha.sched_util_min = tswap32(target_scha->sched_util_min);
> +                scha.sched_util_max = tswap32(target_scha->sched_util_max);
> +            }
> +            unlock_user(target_scha, arg2, 0);
> +            return get_errno(sys_sched_setattr(arg1, &scha, arg3));
> +        }
>       case TARGET_NR_sched_yield:
>           return get_errno(sched_yield());
>       case TARGET_NR_sched_get_priority_max:
> diff --git a/linux-user/syscall_defs.h b/linux-user/syscall_defs.h
> index 0b13975937..310d6ce8ad 100644
> --- a/linux-user/syscall_defs.h
> +++ b/linux-user/syscall_defs.h
> @@ -2914,4 +2914,18 @@ struct target_statx {
>      /* 0x100 */
>   };
>   
> +/* from kernel's include/linux/sched/types.h */
> +struct target_sched_attr {
> +    abi_uint size;
> +    abi_uint sched_policy;
> +    abi_ullong sched_flags;
> +    abi_int sched_nice;
> +    abi_uint sched_priority;
> +    abi_ullong sched_runtime;
> +    abi_ullong sched_deadline;
> +    abi_ullong sched_period;
> +    abi_uint sched_util_min;
> +    abi_uint sched_util_max;
> +};
> +
>   #endif

Reviewed-by: Laurent Vivier <laurent@vivier.eu>

Re: [PATCH v4 2/2] linux-user: call set/getscheduler set/getparam directly

[Laurent Vivier]

Le 05/01/2022 à 05:18, Tonis Tiigi a écrit :
> There seems to be difference in syscall and libc definition of these
> methods and therefore musl does not implement them (1e21e78bf7). Call
> syscall directly to ensure the behavior of the libc of user application,
> not the libc that was used to build QEMU.
> 
> Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
> ---
>   linux-user/syscall.c      | 34 ++++++++++++++++++++++++----------
>   linux-user/syscall_defs.h |  4 ++++
>   2 files changed, 28 insertions(+), 10 deletions(-)
> 
> diff --git a/linux-user/syscall.c b/linux-user/syscall.c
> index 1b8415c8a3..30962155a6 100644
> --- a/linux-user/syscall.c
> +++ b/linux-user/syscall.c
> @@ -359,6 +359,17 @@ _syscall4(int, sys_sched_getattr, pid_t, pid, struct sched_attr *, attr,
>   #define __NR_sys_sched_setattr __NR_sched_setattr
>   _syscall3(int, sys_sched_setattr, pid_t, pid, struct sched_attr *, attr,
>             unsigned int, flags);
> +#define __NR_sys_sched_getscheduler __NR_sched_getscheduler
> +_syscall1(int, sys_sched_getscheduler, pid_t, pid);
> +#define __NR_sys_sched_setscheduler __NR_sched_setscheduler
> +_syscall3(int, sys_sched_setscheduler, pid_t, pid, int, policy,
> +          const struct sched_param *, param);
> +#define __NR_sys_sched_getparam __NR_sched_getparam
> +_syscall2(int, sys_sched_getparam, pid_t, pid,
> +          struct sched_param *, param);
> +#define __NR_sys_sched_setparam __NR_sched_setparam
> +_syscall2(int, sys_sched_setparam, pid_t, pid,
> +          const struct sched_param *, param);
>   #define __NR_sys_getcpu __NR_getcpu
>   _syscall3(int, sys_getcpu, unsigned *, cpu, unsigned *, node, void *, tcache);
>   _syscall4(int, reboot, int, magic1, int, magic2, unsigned int, cmd,
> @@ -10587,30 +10598,32 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
>           return ret;
>       case TARGET_NR_sched_setparam:
>           {
> -            struct sched_param *target_schp;
> +            struct target_sched_param *target_schp;
>               struct sched_param schp;
>   
>               if (arg2 == 0) {
>                   return -TARGET_EINVAL;
>               }
> -            if (!lock_user_struct(VERIFY_READ, target_schp, arg2, 1))
> +            if (!lock_user_struct(VERIFY_READ, target_schp, arg2, 1)) {
>                   return -TARGET_EFAULT;
> +            }
>               schp.sched_priority = tswap32(target_schp->sched_priority);
>               unlock_user_struct(target_schp, arg2, 0);
> -            return get_errno(sched_setparam(arg1, &schp));
> +            return get_errno(sys_sched_setparam(arg1, &schp));
>           }
>       case TARGET_NR_sched_getparam:
>           {
> -            struct sched_param *target_schp;
> +            struct target_sched_param *target_schp;
>               struct sched_param schp;
>   
>               if (arg2 == 0) {
>                   return -TARGET_EINVAL;
>               }
> -            ret = get_errno(sched_getparam(arg1, &schp));
> +            ret = get_errno(sys_sched_getparam(arg1, &schp));
>               if (!is_error(ret)) {
> -                if (!lock_user_struct(VERIFY_WRITE, target_schp, arg2, 0))
> +                if (!lock_user_struct(VERIFY_WRITE, target_schp, arg2, 0)) {
>                       return -TARGET_EFAULT;
> +                }
>                   target_schp->sched_priority = tswap32(schp.sched_priority);
>                   unlock_user_struct(target_schp, arg2, 1);
>               }
> @@ -10618,19 +10631,20 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
>           return ret;
>       case TARGET_NR_sched_setscheduler:
>           {
> -            struct sched_param *target_schp;
> +            struct target_sched_param *target_schp;
>               struct sched_param schp;
>               if (arg3 == 0) {
>                   return -TARGET_EINVAL;
>               }
> -            if (!lock_user_struct(VERIFY_READ, target_schp, arg3, 1))
> +            if (!lock_user_struct(VERIFY_READ, target_schp, arg3, 1)) {
>                   return -TARGET_EFAULT;
> +            }
>               schp.sched_priority = tswap32(target_schp->sched_priority);
>               unlock_user_struct(target_schp, arg3, 0);
> -            return get_errno(sched_setscheduler(arg1, arg2, &schp));
> +            return get_errno(sys_sched_setscheduler(arg1, arg2, &schp));
>           }
>       case TARGET_NR_sched_getscheduler:
> -        return get_errno(sched_getscheduler(arg1));
> +        return get_errno(sys_sched_getscheduler(arg1));
>       case TARGET_NR_sched_getattr:
>           {
>               struct target_sched_attr *target_scha;
> diff --git a/linux-user/syscall_defs.h b/linux-user/syscall_defs.h
> index 310d6ce8ad..28b9fe9a47 100644
> --- a/linux-user/syscall_defs.h
> +++ b/linux-user/syscall_defs.h
> @@ -2928,4 +2928,8 @@ struct target_sched_attr {
>       abi_uint sched_util_max;
>   };
>   
> +struct target_sched_param {
> +    abi_int sched_priority;
> +};
> +
>   #endif

Reviewed-by: Laurent Vivier <laurent@vivier.eu>

Re: [PATCH v4 1/2] linux-user: add sched_getattr support

[Laurent Vivier]

Le 05/01/2022 à 05:18, Tonis Tiigi a écrit :
> These syscalls are not exposed by glibc. The struct type need to be
> redefined as it can't be included directly before
> https://lkml.org/lkml/2020/5/28/810 .
> 
> sched_attr type can grow in future kernel versions. When client sends
> values that QEMU does not understand it will return E2BIG with same
> semantics as old kernel would so client can retry with smaller inputs.
> 
> Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
> ---
>   linux-user/syscall.c      | 123 ++++++++++++++++++++++++++++++++++++++
>   linux-user/syscall_defs.h |  14 +++++
>   2 files changed, 137 insertions(+)
> 
> diff --git a/linux-user/syscall.c b/linux-user/syscall.c
> index 56a3e17183..1b8415c8a3 100644
> --- a/linux-user/syscall.c
> +++ b/linux-user/syscall.c
> @@ -340,6 +340,25 @@ _syscall3(int, sys_sched_getaffinity, pid_t, pid, unsigned int, len,
>   #define __NR_sys_sched_setaffinity __NR_sched_setaffinity
>   _syscall3(int, sys_sched_setaffinity, pid_t, pid, unsigned int, len,
>             unsigned long *, user_mask_ptr);
> +/* sched_attr is not defined in glibc */
> +struct sched_attr {
> +    uint32_t size;
> +    uint32_t sched_policy;
> +    uint64_t sched_flags;
> +    int32_t sched_nice;
> +    uint32_t sched_priority;
> +    uint64_t sched_runtime;
> +    uint64_t sched_deadline;
> +    uint64_t sched_period;
> +    uint32_t sched_util_min;
> +    uint32_t sched_util_max;
> +};
> +#define __NR_sys_sched_getattr __NR_sched_getattr
> +_syscall4(int, sys_sched_getattr, pid_t, pid, struct sched_attr *, attr,
> +          unsigned int, size, unsigned int, flags);
> +#define __NR_sys_sched_setattr __NR_sched_setattr
> +_syscall3(int, sys_sched_setattr, pid_t, pid, struct sched_attr *, attr,
> +          unsigned int, flags);
>   #define __NR_sys_getcpu __NR_getcpu
>   _syscall3(int, sys_getcpu, unsigned *, cpu, unsigned *, node, void *, tcache);
>   _syscall4(int, reboot, int, magic1, int, magic2, unsigned int, cmd,
> @@ -558,6 +577,24 @@ const char *target_strerror(int err)
>       return strerror(target_to_host_errno(err));
>   }
>   
> +static int check_zeroed_user(abi_long addr, size_t ksize, size_t usize)
> +{
> +    int i;
> +    uint8_t b;
> +    if (usize <= ksize) {
> +        return 1;
> +    }
> +    for (i = ksize; i < usize; i++) {
> +        if (get_user_u8(b, addr + i)) {
> +            return -TARGET_EFAULT;
> +        }
> +        if (b != 0) {
> +            return 0;
> +        }
> +    }
> +    return 1;
> +}
> +
>   #define safe_syscall0(type, name) \
>   static type safe_##name(void) \
>   { \
> @@ -10594,6 +10631,92 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
>           }
>       case TARGET_NR_sched_getscheduler:
>           return get_errno(sched_getscheduler(arg1));
> +    case TARGET_NR_sched_getattr:
> +        {
> +            struct target_sched_attr *target_scha;
> +            struct sched_attr scha;
> +            if (arg2 == 0) {
> +                return -TARGET_EINVAL;
> +            }
> +            if (arg3 > sizeof(scha)) {
> +                arg3 = sizeof(scha);
> +            }
> +            ret = get_errno(sys_sched_getattr(arg1, &scha, arg3, arg4));
> +            if (!is_error(ret)) {
> +                target_scha = lock_user(VERIFY_WRITE, arg2, arg3, 0);
> +                if (!target_scha) {
> +                    return -TARGET_EFAULT;
> +                }
> +                target_scha->size = tswap32(scha.size);
> +                target_scha->sched_policy = tswap32(scha.sched_policy);
> +                target_scha->sched_flags = tswap64(scha.sched_flags);
> +                target_scha->sched_nice = tswap32(scha.sched_nice);
> +                target_scha->sched_priority = tswap32(scha.sched_priority);
> +                target_scha->sched_runtime = tswap64(scha.sched_runtime);
> +                target_scha->sched_deadline = tswap64(scha.sched_deadline);
> +                target_scha->sched_period = tswap64(scha.sched_period);
> +                if (scha.size > offsetof(struct sched_attr, sched_util_min)) {
> +                    target_scha->sched_util_min = tswap32(scha.sched_util_min);
> +                    target_scha->sched_util_max = tswap32(scha.sched_util_max);
> +                }
> +                unlock_user(target_scha, arg2, arg3);
> +            }
> +            return ret;
> +        }
> +    case TARGET_NR_sched_setattr:
> +        {
> +            struct target_sched_attr *target_scha;
> +            struct sched_attr scha;
> +            uint32_t size;
> +            int zeroed;
> +            if (arg2 == 0) {
> +                return -TARGET_EINVAL;
> +            }
> +            if (get_user_u32(size, arg2)) {
> +                return -TARGET_EFAULT;
> +            }
> +            if (!size) {
> +                size = offsetof(struct target_sched_attr, sched_util_min);
> +            }
> +            if (size < offsetof(struct target_sched_attr, sched_util_min)) {
> +                if (put_user_u32(sizeof(struct target_sched_attr), arg2)) {
> +                    return -TARGET_EFAULT;
> +                }
> +                return -TARGET_E2BIG;
> +            }
> +
> +            zeroed = check_zeroed_user(arg2, sizeof(struct target_sched_attr), size);
> +            if (zeroed < 0) {
> +                return zeroed;
> +            } else if (zeroed == 0) {
> +                if (put_user_u32(sizeof(struct target_sched_attr), arg2)) {
> +                    return -TARGET_EFAULT;
> +                }
> +                return -TARGET_E2BIG;
> +            }
> +            if (size > sizeof(struct target_sched_attr)) {
> +                size = sizeof(struct target_sched_attr);
> +            }
> +
> +            target_scha = lock_user(VERIFY_READ, arg2, size, 1);
> +            if (!target_scha) {
> +                return -TARGET_EFAULT;
> +            }
> +            scha.size = size;
> +            scha.sched_policy = tswap32(target_scha->sched_policy);
> +            scha.sched_flags = tswap64(target_scha->sched_flags);
> +            scha.sched_nice = tswap32(target_scha->sched_nice);
> +            scha.sched_priority = tswap32(target_scha->sched_priority);
> +            scha.sched_runtime = tswap64(target_scha->sched_runtime);
> +            scha.sched_deadline = tswap64(target_scha->sched_deadline);
> +            scha.sched_period = tswap64(target_scha->sched_period);
> +            if (size > offsetof(struct target_sched_attr, sched_util_min)) {
> +                scha.sched_util_min = tswap32(target_scha->sched_util_min);
> +                scha.sched_util_max = tswap32(target_scha->sched_util_max);
> +            }
> +            unlock_user(target_scha, arg2, 0);
> +            return get_errno(sys_sched_setattr(arg1, &scha, arg3));
> +        }
>       case TARGET_NR_sched_yield:
>           return get_errno(sched_yield());
>       case TARGET_NR_sched_get_priority_max:
> diff --git a/linux-user/syscall_defs.h b/linux-user/syscall_defs.h
> index 0b13975937..310d6ce8ad 100644
> --- a/linux-user/syscall_defs.h
> +++ b/linux-user/syscall_defs.h
> @@ -2914,4 +2914,18 @@ struct target_statx {
>      /* 0x100 */
>   };
>   
> +/* from kernel's include/linux/sched/types.h */
> +struct target_sched_attr {
> +    abi_uint size;
> +    abi_uint sched_policy;
> +    abi_ullong sched_flags;
> +    abi_int sched_nice;
> +    abi_uint sched_priority;
> +    abi_ullong sched_runtime;
> +    abi_ullong sched_deadline;
> +    abi_ullong sched_period;
> +    abi_uint sched_util_min;
> +    abi_uint sched_util_max;
> +};
> +
>   #endif

Applied to my linux-user-for-7.0 branch.

Thanks,
Laurent

Re: [PATCH v4 2/2] linux-user: call set/getscheduler set/getparam directly

[Laurent Vivier]

Le 05/01/2022 à 05:18, Tonis Tiigi a écrit :
> There seems to be difference in syscall and libc definition of these
> methods and therefore musl does not implement them (1e21e78bf7). Call
> syscall directly to ensure the behavior of the libc of user application,
> not the libc that was used to build QEMU.
> 
> Signed-off-by: Tonis Tiigi <tonistiigi@gmail.com>
> ---
>   linux-user/syscall.c      | 34 ++++++++++++++++++++++++----------
>   linux-user/syscall_defs.h |  4 ++++
>   2 files changed, 28 insertions(+), 10 deletions(-)
> 
> diff --git a/linux-user/syscall.c b/linux-user/syscall.c
> index 1b8415c8a3..30962155a6 100644
> --- a/linux-user/syscall.c
> +++ b/linux-user/syscall.c
> @@ -359,6 +359,17 @@ _syscall4(int, sys_sched_getattr, pid_t, pid, struct sched_attr *, attr,
>   #define __NR_sys_sched_setattr __NR_sched_setattr
>   _syscall3(int, sys_sched_setattr, pid_t, pid, struct sched_attr *, attr,
>             unsigned int, flags);
> +#define __NR_sys_sched_getscheduler __NR_sched_getscheduler
> +_syscall1(int, sys_sched_getscheduler, pid_t, pid);
> +#define __NR_sys_sched_setscheduler __NR_sched_setscheduler
> +_syscall3(int, sys_sched_setscheduler, pid_t, pid, int, policy,
> +          const struct sched_param *, param);
> +#define __NR_sys_sched_getparam __NR_sched_getparam
> +_syscall2(int, sys_sched_getparam, pid_t, pid,
> +          struct sched_param *, param);
> +#define __NR_sys_sched_setparam __NR_sched_setparam
> +_syscall2(int, sys_sched_setparam, pid_t, pid,
> +          const struct sched_param *, param);
>   #define __NR_sys_getcpu __NR_getcpu
>   _syscall3(int, sys_getcpu, unsigned *, cpu, unsigned *, node, void *, tcache);
>   _syscall4(int, reboot, int, magic1, int, magic2, unsigned int, cmd,
> @@ -10587,30 +10598,32 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
>           return ret;
>       case TARGET_NR_sched_setparam:
>           {
> -            struct sched_param *target_schp;
> +            struct target_sched_param *target_schp;
>               struct sched_param schp;
>   
>               if (arg2 == 0) {
>                   return -TARGET_EINVAL;
>               }
> -            if (!lock_user_struct(VERIFY_READ, target_schp, arg2, 1))
> +            if (!lock_user_struct(VERIFY_READ, target_schp, arg2, 1)) {
>                   return -TARGET_EFAULT;
> +            }
>               schp.sched_priority = tswap32(target_schp->sched_priority);
>               unlock_user_struct(target_schp, arg2, 0);
> -            return get_errno(sched_setparam(arg1, &schp));
> +            return get_errno(sys_sched_setparam(arg1, &schp));
>           }
>       case TARGET_NR_sched_getparam:
>           {
> -            struct sched_param *target_schp;
> +            struct target_sched_param *target_schp;
>               struct sched_param schp;
>   
>               if (arg2 == 0) {
>                   return -TARGET_EINVAL;
>               }
> -            ret = get_errno(sched_getparam(arg1, &schp));
> +            ret = get_errno(sys_sched_getparam(arg1, &schp));
>               if (!is_error(ret)) {
> -                if (!lock_user_struct(VERIFY_WRITE, target_schp, arg2, 0))
> +                if (!lock_user_struct(VERIFY_WRITE, target_schp, arg2, 0)) {
>                       return -TARGET_EFAULT;
> +                }
>                   target_schp->sched_priority = tswap32(schp.sched_priority);
>                   unlock_user_struct(target_schp, arg2, 1);
>               }
> @@ -10618,19 +10631,20 @@ static abi_long do_syscall1(void *cpu_env, int num, abi_long arg1,
>           return ret;
>       case TARGET_NR_sched_setscheduler:
>           {
> -            struct sched_param *target_schp;
> +            struct target_sched_param *target_schp;
>               struct sched_param schp;
>               if (arg3 == 0) {
>                   return -TARGET_EINVAL;
>               }
> -            if (!lock_user_struct(VERIFY_READ, target_schp, arg3, 1))
> +            if (!lock_user_struct(VERIFY_READ, target_schp, arg3, 1)) {
>                   return -TARGET_EFAULT;
> +            }
>               schp.sched_priority = tswap32(target_schp->sched_priority);
>               unlock_user_struct(target_schp, arg3, 0);
> -            return get_errno(sched_setscheduler(arg1, arg2, &schp));
> +            return get_errno(sys_sched_setscheduler(arg1, arg2, &schp));
>           }
>       case TARGET_NR_sched_getscheduler:
> -        return get_errno(sched_getscheduler(arg1));
> +        return get_errno(sys_sched_getscheduler(arg1));
>       case TARGET_NR_sched_getattr:
>           {
>               struct target_sched_attr *target_scha;
> diff --git a/linux-user/syscall_defs.h b/linux-user/syscall_defs.h
> index 310d6ce8ad..28b9fe9a47 100644
> --- a/linux-user/syscall_defs.h
> +++ b/linux-user/syscall_defs.h
> @@ -2928,4 +2928,8 @@ struct target_sched_attr {
>       abi_uint sched_util_max;
>   };
>   
> +struct target_sched_param {
> +    abi_int sched_priority;
> +};
> +
>   #endif

Applied to my linux-user-for-7.0 branch.

Thanks,
Laurent