[PULL 44/51] bsd-user: Implment madvise(2) to match the linux-user implementation.

[Warner Losh <imp@bsdimp.com>]

From: Karim Taha <kariem.taha2.7@gmail.com>

Signed-off-by: Karim Taha <kariem.taha2.7@gmail.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230925182709.4834-17-kariem.taha2.7@gmail.com>
---
 bsd-user/bsd-mem.h            | 53 +++++++++++++++++++++++++++++++++++
 bsd-user/freebsd/os-syscall.c |  4 +++
 bsd-user/syscall_defs.h       |  2 ++
 3 files changed, 59 insertions(+)

diff --git a/bsd-user/bsd-mem.h b/bsd-user/bsd-mem.h
index 16c22593bfd..b00ab3aed8e 100644
--- a/bsd-user/bsd-mem.h
+++ b/bsd-user/bsd-mem.h
@@ -129,6 +129,59 @@ static inline abi_long do_bsd_munlockall(void)
     return get_errno(munlockall());
 }
 
+/* madvise(2) */
+static inline abi_long do_bsd_madvise(abi_long arg1, abi_long arg2,
+        abi_long arg3)
+{
+    abi_ulong len;
+    int ret = 0;
+    abi_long start = arg1;
+    abi_long len_in = arg2;
+    abi_long advice = arg3;
+
+    if (start & ~TARGET_PAGE_MASK) {
+        return -TARGET_EINVAL;
+    }
+    if (len_in == 0) {
+        return 0;
+    }
+    len = TARGET_PAGE_ALIGN(len_in);
+    if (len == 0 || !guest_range_valid_untagged(start, len)) {
+        return -TARGET_EINVAL;
+    }
+
+    /*
+     * Most advice values are hints, so ignoring and returning success is ok.
+     *
+     * However, some advice values such as MADV_DONTNEED, are not hints and
+     * need to be emulated.
+     *
+     * A straight passthrough for those may not be safe because qemu sometimes
+     * turns private file-backed mappings into anonymous mappings.
+     * If all guest pages have PAGE_PASSTHROUGH set, mappings have the
+     * same semantics for the host as for the guest.
+     *
+     * MADV_DONTNEED is passed through, if possible.
+     * If passthrough isn't possible, we nevertheless (wrongly!) return
+     * success, which is broken but some userspace programs fail to work
+     * otherwise. Completely implementing such emulation is quite complicated
+     * though.
+     */
+    mmap_lock();
+    switch (advice) {
+    case MADV_DONTNEED:
+        if (page_check_range(start, len, PAGE_PASSTHROUGH)) {
+            ret = get_errno(madvise(g2h_untagged(start), len, advice));
+            if (ret == 0) {
+                page_reset_target_data(start, start + len - 1);
+            }
+        }
+    }
+    mmap_unlock();
+
+    return ret;
+}
+
 /* minherit(2) */
 static inline abi_long do_bsd_minherit(abi_long addr, abi_long len,
         abi_long inherit)
diff --git a/bsd-user/freebsd/os-syscall.c b/bsd-user/freebsd/os-syscall.c
index 7a7ae26793f..b8c44cea0ff 100644
--- a/bsd-user/freebsd/os-syscall.c
+++ b/bsd-user/freebsd/os-syscall.c
@@ -831,6 +831,10 @@ static abi_long freebsd_syscall(void *cpu_env, int num, abi_long arg1,
         ret = do_bsd_munlockall();
         break;
 
+    case TARGET_FREEBSD_NR_madvise: /* madvise(2) */
+        ret = do_bsd_madvise(arg1, arg2, arg3);
+        break;
+
     case TARGET_FREEBSD_NR_minherit: /* minherit(2) */
         ret = do_bsd_minherit(arg1, arg2, arg3);
         break;
diff --git a/bsd-user/syscall_defs.h b/bsd-user/syscall_defs.h
index ff692814333..52f84d5dd17 100644
--- a/bsd-user/syscall_defs.h
+++ b/bsd-user/syscall_defs.h
@@ -95,6 +95,8 @@ struct bsd_shm_regions {
 /*
  *  sys/mman.h
  */
+#define TARGET_MADV_DONTNEED            4       /* dont need these pages */
+
 #define TARGET_FREEBSD_MAP_RESERVED0080 0x0080  /* previously misimplemented */
                                                 /* MAP_INHERIT */
 #define TARGET_FREEBSD_MAP_RESERVED0100 0x0100  /* previously unimplemented */
-- 
2.41.0