[PATCH 01/10] linux-user: implement fsmount(2) series of syscalls

[Helge Deller <deller@kernel.org>]

From: Xinhui Yang <cyan@cyano.uk>

This series of syscalls replaces the old mount(2) syscall with a series
of syscalls that operates around a filesystem context. This series of
syscalls is available since Linux 5.2 and glibc 2.36+.

Their users include systemd since v259 and libmount from util-linux, and
possibly other widely used projects.

Preliminary checks are implemented to ensure the validity of the
interface.

Signed-off-by: Xinhui Yang <cyan@cyano.uk>
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@oss.qualcomm.com>
Signed-off-by: Helge Deller <deller@gmx.de>
---
 linux-user/syscall.c | 91 ++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 91 insertions(+)

diff --git a/linux-user/syscall.c b/linux-user/syscall.c
index 7d7a7b489c..c5c2aa80ec 100644
--- a/linux-user/syscall.c
+++ b/linux-user/syscall.c
@@ -14412,6 +14412,97 @@ static abi_long do_syscall1(CPUArchState *cpu_env, int num, abi_long arg1,
         return do_map_shadow_stack(cpu_env, arg1, arg2, arg3);
 #endif
 
+#if defined(TARGET_NR_fsopen)
+    case TARGET_NR_fsopen:
+        {
+            p = lock_user_string(arg1);
+            if (!p) {
+                return -TARGET_EFAULT;
+            }
+            ret = get_errno(fsopen(p, arg2));
+            unlock_user(p, arg1, 0);
+        }
+        return ret;
+    case TARGET_NR_fsconfig:
+        {
+            /*
+             * fsconfig(int, int, char *, void *, int)
+             * NOTE: p4 is nullable and its type might not be a string.
+             */
+            void *p3, *p4;
+            int cmd = (int) arg2;
+            switch (cmd) {
+            case FSCONFIG_SET_BINARY:
+            case FSCONFIG_SET_STRING:
+            case FSCONFIG_SET_PATH:
+            case FSCONFIG_SET_PATH_EMPTY:
+                p3 = lock_user_string(arg3);
+                if (!p3) {
+                    return -TARGET_EFAULT;
+                }
+                if (cmd != FSCONFIG_SET_BINARY) {
+                    /* key and value must be strings. */
+                    p4 = lock_user_string(arg4);
+                } else {
+                    /*
+                     * Otherwise the value must be a raw buffer with its
+                     * length specified in arg5 (aux).
+                     */
+                    p4 = lock_user(VERIFY_READ, arg4, arg5, 1);
+                }
+                if (!p4) {
+                    unlock_user(p3, arg3, 0);
+                    return -TARGET_EFAULT;
+                }
+                ret = get_errno(fsconfig(arg1, arg2, p3, p4, arg5));
+                unlock_user(p3, arg3, 0);
+                unlock_user(p4, arg4, 0);
+                break;
+
+            case FSCONFIG_SET_FLAG:
+            case FSCONFIG_SET_FD:
+                /* arg4 (value) must be NULL. */
+                if (arg4) {
+                    return -TARGET_EFAULT;
+                }
+                p3 = lock_user_string(arg3);
+                if (!p3) {
+                    return -TARGET_EFAULT;
+                }
+                ret = get_errno(fsconfig(arg1, arg2, p3, NULL, arg5));
+                unlock_user(p3, arg3, 0);
+                break;
+            case FSCONFIG_CMD_CREATE:
+            case FSCONFIG_CMD_RECONFIGURE:
+#ifdef FSCONFIG_CMD_CREATE_EXCL
+            /*
+             * FSCONFIG_CMD_CREATE_EXCL is only available since Linux
+             * 6.6. Guarding it to allow building with pre-6.6 headers.
+             */
+            case FSCONFIG_CMD_CREATE_EXCL:
+#endif
+                /* key and value must be NULL, aux must be 0. */
+                if (arg3 || arg4 || arg5) {
+                    return -TARGET_EFAULT;
+                }
+                ret = get_errno(fsconfig(arg1, arg2, NULL, NULL, 0));
+                break;
+            default:
+                return -TARGET_EFAULT;
+            }
+        }
+        return ret;
+    case TARGET_NR_fsmount:
+        ret = get_errno(fsmount(arg1, arg2, arg3));
+        return ret;
+    case TARGET_NR_fspick:
+        {
+            p = lock_user_string(arg2);
+            ret = get_errno(fspick(arg1, p, arg3));
+            unlock_user(p, arg2, 0);
+        }
+        return ret;
+#endif
     default:
         qemu_log_mask(LOG_UNIMP, "Unsupported syscall: %d\n", num);
         return -TARGET_ENOSYS;
-- 
2.54.0