Skip to contentSkip to conversation

9p: Add support for darwin

PATCH Superseded Will Cohen 8 participants 44 messages Feb 6, 2022 – Feb 7, 2022 Arm/Amlogic Meson SoC support

v4 of 9 Superseded Latest is v9 →

lore
Share & tools
lore

Will Cohen

author
#0 lore raw

v4 cover 9p: Add support for darwin

This is a followup to https://lists.nongnu.org/archive/html/qemu-devel/2022-01/msg05993.html,
adding 9p server support for Darwin.

Since v3, the following changes have been made:

- Move XATTR_SIZE_MAX to P9_XATTR_SIZE MAX in 9p.h, and provide explanatory context as preliminary solution
- Add explanatory note surrounding virtio-9p-test with output of pre-patch failing test
- Remove superfluous header guards from file-opt-9p
- Add note about virtfs-proxy-helper being disabled on non-linux for this patch series
- Note radar filed with Apple for missing mknodat syscall
- Replace direct syscall to pthread_fchdir with pthread_fchdir_np, and add check for this function’s presence in meson
- Ensure that d_seekoff is filled using telldir on darwin, and create qemu_dirent_off helper to decide which to access.
- Ensure that [amc]tim.tv_sec are all initialized alongside [amc]tim.tv_nsec in 9p-proxy
- Ensure that all patch email addresses are valid
- Add telldir error handling for dirent on darwin

Keno Fischer (10):
9p: linux: Fix a couple Linux assumptions 9p: Rename 9p-util -> 9p-util-linux
9p: darwin: Handle struct stat(fs) differences 9p: darwin: Handle struct dirent differences 9p: darwin: Ignore O_{NOATIME, DIRECT}
9p: darwin: Move XATTR_SIZE_MAX->P9_XATTR_SIZE_MAX 9p: darwin: *xattr_nofollow implementations 9p: darwin: Compatibility for f/l*xattr 9p: darwin: Implement compatibility for mknodat 9p: darwin: meson: Allow VirtFS on Darwin

Will Cohen (1):
9p: darwin: Adjust assumption on virtio-9p-test

 fsdev/file-op-9p.h                     |  9 ++-
 fsdev/meson.build                      |  1 +
 hw/9pfs/9p-local.c                     | 28 ++++++--
 hw/9pfs/9p-proxy.c                     | 38 ++++++++++-
 hw/9pfs/9p-synth.c                     |  6 ++
 hw/9pfs/9p-util-darwin.c               | 91 ++++++++++++++++++++++++++
 hw/9pfs/{9p-util.c => 9p-util-linux.c} |  7 +-
 hw/9pfs/9p-util.h                      | 38 +++++++++++
 hw/9pfs/9p.c                           | 50 ++++++++++++--
 hw/9pfs/9p.h                           | 11 ++++
 hw/9pfs/codir.c                        |  7 ++
 hw/9pfs/meson.build                    |  3 +-
 include/qemu/xattr.h                   |  4 +-
 meson.build                            | 14 ++--
 tests/qtest/virtio-9p-test.c           |  2 +-
 15 files changed, 285 insertions(+), 24 deletions(-)
 create mode 100644 hw/9pfs/9p-util-darwin.c
 rename hw/9pfs/{9p-util.c => 9p-util-linux.c} (90%)
Signature · 1 line
2.32.0 (Apple Git-132)
author
#1 lore raw

PATCH v4 1/11 9p: linux: Fix a couple Linux assumptions

From: Keno Fischer <[email protected]>

- Guard Linux only headers.
- Add qemu/statfs.h header to abstract over the which headers are needed for struct statfs
- Define ENOATTR only if not only defined (it's defined in system headers on Darwin).

[Michael Roitzsch: - Rebase for NixOS]

While it might at first appear that fsdev/virtfs-proxy-header.c would need similar adjustment for darwin as file-op-9p here, a later patch in this series disables virtfs-proxy-helper for non-Linux. Allowing virtfs-proxy-helper on darwin could potentially be an additional optimization later.

[Will Cohen: - Fix headers for Alpine
- Integrate statfs.h back into file-op-9p.h
- Remove superfluous header guards from file-opt-9p
- Add note about virtfs-proxy-helper being disabled on non-Linux for this patch series]

---
 fsdev/file-op-9p.h   | 9 ++++++++-
 hw/9pfs/9p-local.c   | 2 ++
 hw/9pfs/9p.c         | 4 ++++
 include/qemu/xattr.h | 4 +++-
 4 files changed, 17 insertions(+), 2 deletions(-)
fsdev/file-op-9p.h +8 -1
index 8fd89f0447..4997677460 100644
@@ -16,10 +16,17 @@
 
 #include <dirent.h>
 #include <utime.h>
-#include <sys/vfs.h>
 #include "qemu-fsdev-throttle.h"
 #include "p9array.h"
 
+#ifdef CONFIG_LINUX
+# include <sys/vfs.h>
+#endif
+#ifdef CONFIG_DARWIN
+# include <sys/param.h>
+# include <sys/mount.h>
+#endif
+
 #define SM_LOCAL_MODE_BITS    0600
 #define SM_LOCAL_DIR_MODE_BITS    0700
 
hw/9pfs/9p-local.c +2 -0
index 210d9e7705..1a5e3eed73 100644
@@ -32,10 +32,12 @@
 #include "qemu/error-report.h"
 #include "qemu/option.h"
 #include <libgen.h>
+#ifdef CONFIG_LINUX
 #include <linux/fs.h>
 #ifdef CONFIG_LINUX_MAGIC_H
 #include <linux/magic.h>
 #endif
+#endif
 #include <sys/ioctl.h>
 
 #ifndef XFS_SUPER_MAGIC
hw/9pfs/9p.c +4 -0
index 15b3f4d385..9c63e14b28 100644
@@ -32,7 +32,11 @@
 #include "migration/blocker.h"
 #include "qemu/xxhash.h"
 #include <math.h>
+#ifdef CONFIG_LINUX
 #include <linux/limits.h>
+#else
+#include <limits.h>
+#endif
 
 int open_fd_hw;
 int total_open_fd;
include/qemu/xattr.h +3 -1
index a83fe8e749..f1d0f7be74 100644
@@ -22,7 +22,9 @@
 #ifdef CONFIG_LIBATTR
 #  include <attr/xattr.h>
 #else
-#  define ENOATTR ENODATA
+#  if !defined(ENOATTR)
+#    define ENOATTR ENODATA
+#  endif
 #  include <sys/xattr.h>
 #endif
 
Signature · 1 line
2.32.0 (Apple Git-132)
Reviewed-by Philippe Mathieu-Daudé on v4 1/11
#2 lore raw

Re: [PATCH v4 01/11] 9p: linux: Fix a couple Linux assumptions

Quoted 23 lines · Will Cohen

From: Keno Fischer <[email protected]>

- Guard Linux only headers.
- Add qemu/statfs.h header to abstract over the which headers are needed for struct statfs
- Define ENOATTR only if not only defined (it's defined in system headers on Darwin).

[Michael Roitzsch: - Rebase for NixOS]

While it might at first appear that fsdev/virtfs-proxy-header.c would need similar adjustment for darwin as file-op-9p here, a later patch in this series disables virtfs-proxy-helper for non-Linux. Allowing virtfs-proxy-helper on darwin could potentially be an additional optimization later.

[Will Cohen: - Fix headers for Alpine
- Integrate statfs.h back into file-op-9p.h
- Remove superfluous header guards from file-opt-9p
- Add note about virtfs-proxy-helper being disabled on non-Linux for this patch series]

Quoted 7 lines
---
  fsdev/file-op-9p.h   | 9 ++++++++-
  hw/9pfs/9p-local.c   | 2 ++
  hw/9pfs/9p.c         | 4 ++++
  include/qemu/xattr.h | 4 +++-
  4 files changed, 17 insertions(+), 2 deletions(-)
Reviewed-by on v4 1/11
#3 lore raw

Re: [PATCH v4 01/11] 9p: linux: Fix a couple Linux assumptions

In reply to Will Cohen
Quoted 18 lines · On Sun, 6 Feb 2022 15:07:09 -0500 Will Cohen <[email protected]> wrote:

From: Keno Fischer <[email protected]>

- Guard Linux only headers.
- Add qemu/statfs.h header to abstract over the which headers are needed for struct statfs
- Define ENOATTR only if not only defined (it's defined in system headers on Darwin).

[Michael Roitzsch: - Rebase for NixOS]

While it might at first appear that fsdev/virtfs-proxy-header.c would need similar adjustment for darwin as file-op-9p here, a later patch in this series disables virtfs-proxy-helper for non-Linux. Allowing virtfs-proxy-helper on darwin could potentially be an additional optimization later.

Quoted 7 lines

[Will Cohen: - Fix headers for Alpine
- Integrate statfs.h back into file-op-9p.h
- Remove superfluous header guards from file-opt-9p
- Add note about virtfs-proxy-helper being disabled on non-Linux for this patch series]

---

Quoted 77 lines
 fsdev/file-op-9p.h   | 9 ++++++++-
 hw/9pfs/9p-local.c   | 2 ++
 hw/9pfs/9p.c         | 4 ++++
 include/qemu/xattr.h | 4 +++-
 4 files changed, 17 insertions(+), 2 deletions(-)
fsdev/file-op-9p.h +8 -1
index 8fd89f0447..4997677460 100644
@@ -16,10 +16,17 @@
 
 #include <dirent.h>
 #include <utime.h>
-#include <sys/vfs.h>
 #include "qemu-fsdev-throttle.h"
 #include "p9array.h"
 
+#ifdef CONFIG_LINUX
+# include <sys/vfs.h>
+#endif
+#ifdef CONFIG_DARWIN
+# include <sys/param.h>
+# include <sys/mount.h>
+#endif
+
 #define SM_LOCAL_MODE_BITS    0600
 #define SM_LOCAL_DIR_MODE_BITS    0700
 
hw/9pfs/9p-local.c +2 -0
index 210d9e7705..1a5e3eed73 100644
@@ -32,10 +32,12 @@
 #include "qemu/error-report.h"
 #include "qemu/option.h"
 #include <libgen.h>
+#ifdef CONFIG_LINUX
 #include <linux/fs.h>
 #ifdef CONFIG_LINUX_MAGIC_H
 #include <linux/magic.h>
 #endif
+#endif
 #include <sys/ioctl.h>
 
 #ifndef XFS_SUPER_MAGIC
hw/9pfs/9p.c +4 -0
index 15b3f4d385..9c63e14b28 100644
@@ -32,7 +32,11 @@
 #include "migration/blocker.h"
 #include "qemu/xxhash.h"
 #include <math.h>
+#ifdef CONFIG_LINUX
 #include <linux/limits.h>
+#else
+#include <limits.h>
+#endif
 
 int open_fd_hw;
 int total_open_fd;
include/qemu/xattr.h +3 -1
index a83fe8e749..f1d0f7be74 100644
@@ -22,7 +22,9 @@
 #ifdef CONFIG_LIBATTR
 #  include <attr/xattr.h>
 #else
-#  define ENOATTR ENODATA
+#  if !defined(ENOATTR)
+#    define ENOATTR ENODATA
+#  endif
 #  include <sys/xattr.h>
 #endif
 
author
#4 lore raw

PATCH v4 2/11 9p: Rename 9p-util -> 9p-util-linux

In reply to Will Cohen

From: Keno Fischer <[email protected]>

The current file only has the Linux versions of these functions. Rename the file accordingly and update the Makefile to only build it on Linux. A Darwin version of these will follow later in the series.

[Michael Roitzsch: - Rebase for NixOS]
---
 hw/9pfs/{9p-util.c => 9p-util-linux.c} | 2 +-
 hw/9pfs/meson.build                    | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
 rename hw/9pfs/{9p-util.c => 9p-util-linux.c} (97%)
hw/9pfs/9p-util-linux.c renamed from hw/9pfs/9p-util.c +1 -1
similarity index 97%
rename from hw/9pfs/9p-util.c
rename to hw/9pfs/9p-util-linux.c
index 3221d9b498..398614a5d0 100644
@@ -1,5 +1,5 @@
 /*
- * 9p utilities
+ * 9p utilities (Linux Implementation)
  *
  * Copyright IBM, Corp. 2017
  *
hw/9pfs/meson.build +1 -1
index 99be5d9119..1b28e70040 100644
@@ -4,7 +4,6 @@ fs_ss.add(files(
   '9p-posix-acl.c',
   '9p-proxy.c',
   '9p-synth.c',
-  '9p-util.c',
   '9p-xattr-user.c',
   '9p-xattr.c',
   '9p.c',
@@ -14,6 +13,7 @@ fs_ss.add(files(
   'coth.c',
   'coxattr.c',
 ))
+fs_ss.add(when: 'CONFIG_LINUX', if_true: files('9p-util-linux.c'))
 fs_ss.add(when: 'CONFIG_XEN', if_true: files('xen-9p-backend.c'))
 softmmu_ss.add_all(when: 'CONFIG_FSDEV_9P', if_true: fs_ss)
 
Signature · 1 line
2.32.0 (Apple Git-132)
Reviewed-by Philippe Mathieu-Daudé on v4 2/11
#5 lore raw

Re: [PATCH v4 02/11] 9p: Rename 9p-util -> 9p-util-linux

Quoted 12 lines · Will Cohen

From: Keno Fischer <[email protected]>

The current file only has the Linux versions of these functions. Rename the file accordingly and update the Makefile to only build it on Linux. A Darwin version of these will follow later in the series.

[Michael Roitzsch: - Rebase for NixOS]
Quoted 5 lines
---
  hw/9pfs/{9p-util.c => 9p-util-linux.c} | 2 +-
  hw/9pfs/meson.build                    | 2 +-
  2 files changed, 2 insertions(+), 2 deletions(-)

rename hw/9pfs/{9p-util.c => 9p-util-linux.c} (97%)

author
#6 lore raw

PATCH v4 3/11 9p: darwin: Handle struct stat(fs) differences

In reply to Will Cohen

From: Keno Fischer <[email protected]>

[Will Cohen: - Note lack of f_namelen and f_frsize on Darwin
- Ensure that tv_sec and tv_nsec are both initialized for Darwin and non-Darwin]

---
 hw/9pfs/9p-proxy.c | 22 ++++++++++++++++++++--
 hw/9pfs/9p-synth.c |  2 ++
 hw/9pfs/9p.c       | 16 ++++++++++++++--
 3 files changed, 36 insertions(+), 4 deletions(-)
hw/9pfs/9p-proxy.c +20 -2
index 09bd9f1464..b1664080d8 100644
@@ -123,10 +123,16 @@ static void prstatfs_to_statfs(struct statfs *stfs, ProxyStatFS *prstfs)
     stfs->f_bavail = prstfs->f_bavail;
     stfs->f_files = prstfs->f_files;
     stfs->f_ffree = prstfs->f_ffree;
+#ifdef CONFIG_DARWIN
+    /* f_namelen and f_frsize do not exist on Darwin */
+    stfs->f_fsid.val[0] = prstfs->f_fsid[0] & 0xFFFFFFFFU;
+    stfs->f_fsid.val[1] = prstfs->f_fsid[1] >> 32 & 0xFFFFFFFFU;
+#else
     stfs->f_fsid.__val[0] = prstfs->f_fsid[0] & 0xFFFFFFFFU;
     stfs->f_fsid.__val[1] = prstfs->f_fsid[1] >> 32 & 0xFFFFFFFFU;
     stfs->f_namelen = prstfs->f_namelen;
     stfs->f_frsize = prstfs->f_frsize;
+#endif
 }
 
 /* Converts proxy_stat structure to VFS stat structure */
@@ -143,12 +149,24 @@ static void prstat_to_stat(struct stat *stbuf, ProxyStat *prstat)
    stbuf->st_size = prstat->st_size;
    stbuf->st_blksize = prstat->st_blksize;
    stbuf->st_blocks = prstat->st_blocks;
+   stbuf->st_atime = prstat->st_atim_sec;
+   stbuf->st_mtime = prstat->st_mtim_sec;
+   stbuf->st_ctime = prstat->st_ctim_sec;
+#ifdef CONFIG_DARWIN
+   stbuf->st_atimespec.tv_sec = prstat->st_atim_sec;
+   stbuf->st_mtimespec.tv_sec = prstat->st_mtim_sec;
+   stbuf->st_ctimespec.tv_sec = prstat->st_ctim_sec;
+   stbuf->st_atimespec.tv_nsec = prstat->st_atim_nsec;
+   stbuf->st_mtimespec.tv_nsec = prstat->st_mtim_nsec;
+   stbuf->st_ctimespec.tv_nsec = prstat->st_ctim_nsec;
+#else
    stbuf->st_atim.tv_sec = prstat->st_atim_sec;
+   stbuf->st_mtim.tv_sec = prstat->st_mtim_sec;
+   stbuf->st_ctim.tv_sec = prstat->st_ctim_sec;
    stbuf->st_atim.tv_nsec = prstat->st_atim_nsec;
-   stbuf->st_mtime = prstat->st_mtim_sec;
    stbuf->st_mtim.tv_nsec = prstat->st_mtim_nsec;
-   stbuf->st_ctime = prstat->st_ctim_sec;
    stbuf->st_ctim.tv_nsec = prstat->st_ctim_nsec;
+#endif
 }
 
 /*
hw/9pfs/9p-synth.c +2 -0
index b38088e066..4a4a776d06 100644
@@ -427,7 +427,9 @@ static int synth_statfs(FsContext *s, V9fsPath *fs_path,
     stbuf->f_bsize = 512;
     stbuf->f_blocks = 0;
     stbuf->f_files = synth_node_count;
+#ifndef CONFIG_DARWIN
     stbuf->f_namelen = NAME_MAX;
+#endif
     return 0;
 }
 
hw/9pfs/9p.c +14 -2
index 9c63e14b28..1563d7b7c6 100644
@@ -1313,11 +1313,17 @@ static int stat_to_v9stat_dotl(V9fsPDU *pdu, const struct stat *stbuf,
     v9lstat->st_blksize = stat_to_iounit(pdu, stbuf);
     v9lstat->st_blocks = stbuf->st_blocks;
     v9lstat->st_atime_sec = stbuf->st_atime;
-    v9lstat->st_atime_nsec = stbuf->st_atim.tv_nsec;
     v9lstat->st_mtime_sec = stbuf->st_mtime;
-    v9lstat->st_mtime_nsec = stbuf->st_mtim.tv_nsec;
     v9lstat->st_ctime_sec = stbuf->st_ctime;
+#ifdef CONFIG_DARWIN
+    v9lstat->st_atime_nsec = stbuf->st_atimespec.tv_nsec;
+    v9lstat->st_mtime_nsec = stbuf->st_mtimespec.tv_nsec;
+    v9lstat->st_ctime_nsec = stbuf->st_ctimespec.tv_nsec;
+#else
+    v9lstat->st_atime_nsec = stbuf->st_atim.tv_nsec;
+    v9lstat->st_mtime_nsec = stbuf->st_mtim.tv_nsec;
     v9lstat->st_ctime_nsec = stbuf->st_ctim.tv_nsec;
+#endif
     /* Currently we only support BASIC fields in stat */
     v9lstat->st_result_mask = P9_STATS_BASIC;
 
@@ -3519,9 +3525,15 @@ static int v9fs_fill_statfs(V9fsState *s, V9fsPDU *pdu, struct statfs *stbuf)
     f_bavail = stbuf->f_bavail / bsize_factor;
     f_files  = stbuf->f_files;
     f_ffree  = stbuf->f_ffree;
+#ifdef CONFIG_DARWIN
+    fsid_val = (unsigned int)stbuf->f_fsid.val[0] |
+               (unsigned long long)stbuf->f_fsid.val[1] << 32;
+    f_namelen = NAME_MAX;
+#else
     fsid_val = (unsigned int) stbuf->f_fsid.__val[0] |
                (unsigned long long)stbuf->f_fsid.__val[1] << 32;
     f_namelen = stbuf->f_namelen;
+#endif
 
     return pdu_marshal(pdu, offset, "ddqqqqqqd",
                        f_type, f_bsize, f_blocks, f_bfree,
Signature · 1 line
2.32.0 (Apple Git-132)
author
#7 lore raw

PATCH v4 4/11 9p: darwin: Handle struct dirent differences

In reply to Will Cohen

From: Keno Fischer <[email protected]>

On darwin d_seekoff exists, but is optional and does not seem to be commonly used by file systems. Use telldir instead to obtain the seek offset and inject it into d_seekoff, and create a qemu_dirent_off helper to call it appropriately when appropriate.

[Michael Roitzsch: - Rebase for NixOS]

[Will Cohen: - Adjust to pass testing
- Ensure that d_seekoff is filled using telldir on darwin, and create qemu_dirent_off helper to decide which to access]
[Fabian Franz: - Add telldir error handling for darwin] [Will Cohen: - Ensure that telldir error handling uses
signed int]

---
 hw/9pfs/9p-local.c |  9 +++++++++
 hw/9pfs/9p-proxy.c | 16 +++++++++++++++-
 hw/9pfs/9p-synth.c |  4 ++++
 hw/9pfs/9p-util.h  | 17 +++++++++++++++++
 hw/9pfs/9p.c       | 15 +++++++++++++--
 hw/9pfs/codir.c    |  7 +++++++
 6 files changed, 65 insertions(+), 3 deletions(-)
hw/9pfs/9p-local.c +9 -0
index 1a5e3eed73..7137a28109 100644
@@ -559,6 +559,15 @@ static struct dirent *local_readdir(FsContext *ctx, V9fsFidOpenState *fs)
 
 again:
     entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
     if (!entry) {
         return NULL;
     }
hw/9pfs/9p-proxy.c +15 -1
index b1664080d8..8b4b5cf7dc 100644
@@ -706,7 +706,21 @@ static off_t proxy_telldir(FsContext *ctx, V9fsFidOpenState *fs)
 
 static struct dirent *proxy_readdir(FsContext *ctx, V9fsFidOpenState *fs)
 {
-    return readdir(fs->dir.stream);
+    struct dirent *entry;
+    entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    if (!entry) {
+        return NULL;
+    }
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
+    return entry;
 }
 
 static void proxy_seekdir(FsContext *ctx, V9fsFidOpenState *fs, off_t off)
hw/9pfs/9p-synth.c +4 -0
index 4a4a776d06..e264a03eef 100644
@@ -222,7 +222,11 @@ static void synth_direntry(V9fsSynthNode *node,
 {
     strcpy(entry->d_name, node->name);
     entry->d_ino = node->attr->inode;
+#ifdef CONFIG_DARWIN
+    entry->d_seekoff = off + 1;
+#else
     entry->d_off = off + 1;
+#endif
 }
 
 static struct dirent *synth_get_dentry(V9fsSynthNode *dir,
hw/9pfs/9p-util.h +17 -0
index 546f46dc7d..accbec9987 100644
@@ -79,3 +79,20 @@ ssize_t fremovexattrat_nofollow(int dirfd, const char *filename,
                                 const char *name);
 
 #endif
+
+
+/**
+ * Darwin has d_seekoff, which appears to function similarly to d_off.
+ * However, it does not appear to be supported on all file systems,
+ * so ensure it is manually injected earlier and call here when
+ * needed.
+ */
+
+inline off_t qemu_dirent_off(struct dirent *dent)
+{
+#ifdef CONFIG_DARWIN
+    return dent->d_seekoff;
+#else
+    return dent->d_off;
+#endif
+}
hw/9pfs/9p.c +13 -2
index 1563d7b7c6..cf694da354 100644
@@ -27,6 +27,7 @@
 #include "virtio-9p.h"
 #include "fsdev/qemu-fsdev.h"
 #include "9p-xattr.h"
+#include "9p-util.h"
 #include "coth.h"
 #include "trace.h"
 #include "migration/blocker.h"
@@ -2281,7 +2282,11 @@ static int coroutine_fn v9fs_do_readdir_with_stat(V9fsPDU *pdu,
         count += len;
         v9fs_stat_free(&v9stat);
         v9fs_path_free(&path);
-        saved_dir_pos = dent->d_off;
+        saved_dir_pos = qemu_dirent_off(dent);
+        if (saved_dir_pos < 0) {
+            err = saved_dir_pos;
+            break;
+        }
     }
 
     v9fs_readdir_unlock(&fidp->fs.dir);
@@ -2420,6 +2425,7 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU *pdu, V9fsFidState *fidp,
     V9fsString name;
     int len, err = 0;
     int32_t count = 0;
+    off_t off;
     struct dirent *dent;
     struct stat *st;
     struct V9fsDirEnt *entries = NULL;
@@ -2480,12 +2486,17 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU *pdu, V9fsFidState *fidp,
             qid.version = 0;
         }
 
+        off = qemu_dirent_off(dent);
+        if (off < 0) {
+            err = off;
+            break;
+        }
         v9fs_string_init(&name);
         v9fs_string_sprintf(&name, "%s", dent->d_name);
 
         /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
         len = pdu_marshal(pdu, 11 + count, "Qqbs",
-                          &qid, dent->d_off,
+                          &qid, off,
                           dent->d_type, &name);
 
         v9fs_string_free(&name);
hw/9pfs/codir.c +7 -0
index 032cce04c4..fac6759a64 100644
@@ -167,7 +167,14 @@ static int do_readdir_many(V9fsPDU *pdu, V9fsFidState *fidp,
         }
 
         size += len;
+        /* This conditional statement is identical in
+         * function to qemu_dirent_off, described in 9p-util.h,
+         * since that header cannot be included here. */
+#ifdef CONFIG_DARWIN
+        saved_dir_pos = dent->d_seekoff;
+#else
         saved_dir_pos = dent->d_off;
+#endif
     }
 
     /* restore (last) saved position */
Signature · 1 line
2.32.0 (Apple Git-132)
#8 lore raw

Re: [PATCH v4 04/11] 9p: darwin: Handle struct dirent differences

Comments inline:

hw/9pfs/9p-local.c +0 -0

hw/9pfs/9p-local.c static struct dirent *local_readdir(FsContext *ctx, · line 565

Quoted 11 lines

index 1a5e3eed73..7137a28109 100644

hw/9pfs/9p-local.c +3 -0
@@ -559,6 +559,15 @@ static struct dirent *local_readdir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 again:
     entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    int td;
+    td = telldir(fs->dir.stream);

Maybe call this „off“?

Quoted 10 lines
hw/9pfs/9p-local.c +6 -0
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
     if (!entry) {
         return NULL;
     }

This needs to be before the #ifdef!

Quoted 60 lines
hw/9pfs/9p-proxy.c +15 -1
index b1664080d8..8b4b5cf7dc 100644
@@ -706,7 +706,21 @@ static off_t proxy_telldir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 static struct dirent *proxy_readdir(FsContext *ctx, V9fsFidOpenState *fs)
 {
-    return readdir(fs->dir.stream);
+    struct dirent *entry;
+    entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    if (!entry) {
+        return NULL;
+    }
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
+    return entry;
 }
 

static void proxy_seekdir(FsContext *ctx, V9fsFidOpenState *fs, off_t off)

hw/9pfs/9p-synth.c +4 -0
index 4a4a776d06..e264a03eef 100644
@@ -222,7 +222,11 @@ static void synth_direntry(V9fsSynthNode *node,
 {
     strcpy(entry->d_name, node->name);
     entry->d_ino = node->attr->inode;
+#ifdef CONFIG_DARWIN
+    entry->d_seekoff = off + 1;
+#else
     entry->d_off = off + 1;
+#endif
 }
 
 static struct dirent *synth_get_dentry(V9fsSynthNode *dir,
hw/9pfs/9p-util.h +7 -0
index 546f46dc7d..accbec9987 100644
@@ -79,3 +79,20 @@ ssize_t fremovexattrat_nofollow(int dirfd, const char
 *filename,
                                 const char *name);
 
 #endif
+
+
+/**
+ * Darwin has d_seekoff, which appears to function similarly to d_off.
+ * However, it does not appear to be supported on all file systems,
+ * so ensure it is manually injected earlier and call here when
+ * needed.

hw/9pfs/9p-util.h ssize_t fremovexattrat_nofollow(int dirfd, const char

Quoted 10 lines
hw/9pfs/9p-util.h +10 -0
+ */
+
+inline off_t qemu_dirent_off(struct dirent *dent)
+{
+#ifdef CONFIG_DARWIN
+    return dent->d_seekoff;
+#else
+    return dent->d_off;
+#endif
+}

Are we sure we want a helper for two times the same ifdef? Deferring to maintainers here however.

hw/9pfs/9p.c +0 -0
Quoted 12 lines

index 1563d7b7c6..cf694da354 100644

hw/9pfs/9p.c +1 -0
@@ -27,6 +27,7 @@
 #include "virtio-9p.h"
 #include "fsdev/qemu-fsdev.h"
 #include "9p-xattr.h"
+#include "9p-util.h"
 #include "coth.h"
 #include "trace.h"
 #include "migration/blocker.h"

hw/9pfs/9p.c static int coroutine_fn · line 2290

Quoted 10 lines
hw/9pfs/9p.c +5 -1
@@ -2281,7 +2282,11 @@ static int coroutine_fn
 v9fs_do_readdir_with_stat(V9fsPDU *pdu,
         count += len;
         v9fs_stat_free(&v9stat);
         v9fs_path_free(&path);
-        saved_dir_pos = dent->d_off;
+        saved_dir_pos = qemu_dirent_off(dent);
+        if (saved_dir_pos < 0) {
+            err = saved_dir_pos;
+            break;
+        }

Do we still need this error-handling? I had removed it in my interdiff patch.

Quoted 14 lines
hw/9pfs/9p.c +0 -0
     }
 
     v9fs_readdir_unlock(&fidp->fs.dir);
hw/9pfs/9p.c +1 -0
@@ -2420,6 +2425,7 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp,
     V9fsString name;
     int len, err = 0;
     int32_t count = 0;
+    off_t off;
     struct dirent *dent;
     struct stat *st;
     struct V9fsDirEnt *entries = NULL;

hw/9pfs/9p.c static int coroutine_fn v9fs_do_readdir(V9fsPDU · line 2494

Quoted 9 lines
hw/9pfs/9p.c +5 -0
@@ -2480,12 +2486,17 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp,
             qid.version = 0;
         }
 
+        off = qemu_dirent_off(dent);
+        if (off < 0) {
+            err = off;
+            break;
+        }

Same here - if this can never fail, why add the error handling?

Quoted 34 lines
hw/9pfs/9p.c +1 -1
         v9fs_string_init(&name);
         v9fs_string_sprintf(&name, "%s", dent->d_name);
 
         /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
         len = pdu_marshal(pdu, 11 + count, "Qqbs",
-                          &qid, dent->d_off,
+                          &qid, off,
                           dent->d_type, &name);
 
         v9fs_string_free(&name);
hw/9pfs/codir.c +7 -0
index 032cce04c4..fac6759a64 100644
@@ -167,7 +167,14 @@ static int do_readdir_many(V9fsPDU *pdu, V9fsFidState
 *fidp,
         }
 
         size += len;
+        /* This conditional statement is identical in
+         * function to qemu_dirent_off, described in 9p-util.h,
+         * since that header cannot be included here. */
+#ifdef CONFIG_DARWIN
+        saved_dir_pos = dent->d_seekoff;
+#else
         saved_dir_pos = dent->d_off;
+#endif
     }
 
     /* restore (last) saved position */

--
2.32.0 (Apple Git-132)

author
#9 lore raw

Re: [PATCH v4 04/11] 9p: darwin: Handle struct dirent differences

Quoted 9 lines · Fabian Franz

Comments inline:

hw/9pfs/9p-local.c +0 -0
Quoted 6 lines

index 1a5e3eed73..7137a28109 100644

hw/9pfs/9p-local.c +0 -0
@@ -559,6 +559,15 @@ static struct dirent *local_readdir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
Quoted 9 lines

hw/9pfs/9p-local.c static struct dirent *local_readdir(FsContext *ctx, · line 565

Quoted 5 lines
hw/9pfs/9p-local.c +3 -0
 again:
     entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    int td;
+    td = telldir(fs->dir.stream);

Maybe call this „off“?

Yes, off is better. Will adjust for v5.

Quoted 12 lines
Quoted 10 lines
hw/9pfs/9p-local.c +6 -0
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
     if (!entry) {
         return NULL;
     }
Quoted 2 lines

This needs to be before the #ifdef!

Good catch, will adjust for v5. I moved it around twice and forgot to put it in the right place.

Quoted 69 lines
Quoted 69 lines
hw/9pfs/9p-proxy.c +15 -1
index b1664080d8..8b4b5cf7dc 100644
@@ -706,7 +706,21 @@ static off_t proxy_telldir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 static struct dirent *proxy_readdir(FsContext *ctx, V9fsFidOpenState *fs)
 {
-    return readdir(fs->dir.stream);
+    struct dirent *entry;
+    entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    if (!entry) {
+        return NULL;
+    }
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
+    return entry;
 }
 

static void proxy_seekdir(FsContext *ctx, V9fsFidOpenState *fs, off_t
off)

hw/9pfs/9p-synth.c +4 -0
index 4a4a776d06..e264a03eef 100644
@@ -222,7 +222,11 @@ static void synth_direntry(V9fsSynthNode *node,
 {
     strcpy(entry->d_name, node->name);
     entry->d_ino = node->attr->inode;
+#ifdef CONFIG_DARWIN
+    entry->d_seekoff = off + 1;
+#else
     entry->d_off = off + 1;
+#endif
 }
 
 static struct dirent *synth_get_dentry(V9fsSynthNode *dir,
hw/9pfs/9p-util.h +13 -0
index 546f46dc7d..accbec9987 100644
@@ -79,3 +79,20 @@ ssize_t fremovexattrat_nofollow(int dirfd, const char
 *filename,
                                 const char *name);
 
 #endif
+
+
+/**
+ * Darwin has d_seekoff, which appears to function similarly to d_off.
+ * However, it does not appear to be supported on all file systems,
+ * so ensure it is manually injected earlier and call here when
+ * needed.
+ */
+
+inline off_t qemu_dirent_off(struct dirent *dent)
+{
+#ifdef CONFIG_DARWIN
+    return dent->d_seekoff;
Quoted 9 lines

hw/9pfs/9p-util.h

Quoted 4 lines
hw/9pfs/9p-util.h +4 -0
+#else
+    return dent->d_off;
+#endif
+}

Are we sure we want a helper for two times the same ifdef? Deferring to maintainers here however.

Either way works for me too -- my current inclination is to leave it this way (as originally suggested by the maintainers), if for no other reason than that it allows the one comment to be referenced in the case of both uses.

Quoted 20 lines
hw/9pfs/9p.c +0 -0
Quoted 18 lines

index 1563d7b7c6..cf694da354 100644

hw/9pfs/9p.c +2 -1
@@ -27,6 +27,7 @@
 #include "virtio-9p.h"
 #include "fsdev/qemu-fsdev.h"
 #include "9p-xattr.h"
+#include "9p-util.h"
 #include "coth.h"
 #include "trace.h"
 #include "migration/blocker.h"
@@ -2281,7 +2282,11 @@ static int coroutine_fn
 v9fs_do_readdir_with_stat(V9fsPDU *pdu,
         count += len;
         v9fs_stat_free(&v9stat);
         v9fs_path_free(&path);
-        saved_dir_pos = dent->d_off;
+        saved_dir_pos = qemu_dirent_off(dent);
Quoted 9 lines

hw/9pfs/9p.c static int coroutine_fn · line 2290

Quoted 4 lines
hw/9pfs/9p.c +4 -0
+        if (saved_dir_pos < 0) {
+            err = saved_dir_pos;
+            break;
+        }

Do we still need this error-handling? I had removed it in my interdiff patch.

That's correct, it in fact can be removed. d_seekoff yields a __uint64_t (
https://developer.apple.com/documentation/kernel/direntry/1415494-d_seekoff?language=objc).
Will adjust for v5.

Quoted 18 lines
Quoted 18 lines
hw/9pfs/9p.c +0 -0
     }
 
     v9fs_readdir_unlock(&fidp->fs.dir);
hw/9pfs/9p.c +1 -0
@@ -2420,6 +2425,7 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp,
     V9fsString name;
     int len, err = 0;
     int32_t count = 0;
+    off_t off;
     struct dirent *dent;
     struct stat *st;
     struct V9fsDirEnt *entries = NULL;
hw/9pfs/9p.c +0 -0
@@ -2480,12 +2486,17 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp,
             qid.version = 0;
         }
 
Quoted 9 lines

hw/9pfs/9p.c static int coroutine_fn v9fs_do_readdir(V9fsPDU · line 2494

Quoted 5 lines
hw/9pfs/9p.c +5 -0
+        off = qemu_dirent_off(dent);
+        if (off < 0) {
+            err = off;
+            break;
+        }

Same here - if this can never fail, why add the error handling?

See above.

Quoted 36 lines
Quoted 36 lines
hw/9pfs/9p.c +1 -1
         v9fs_string_init(&name);
         v9fs_string_sprintf(&name, "%s", dent->d_name);
 
         /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
         len = pdu_marshal(pdu, 11 + count, "Qqbs",
-                          &qid, dent->d_off,
+                          &qid, off,
                           dent->d_type, &name);
 
         v9fs_string_free(&name);
hw/9pfs/codir.c +7 -0
index 032cce04c4..fac6759a64 100644
@@ -167,7 +167,14 @@ static int do_readdir_many(V9fsPDU *pdu,
 V9fsFidState *fidp,
         }
 
         size += len;
+        /* This conditional statement is identical in
+         * function to qemu_dirent_off, described in 9p-util.h,
+         * since that header cannot be included here. */
+#ifdef CONFIG_DARWIN
+        saved_dir_pos = dent->d_seekoff;
+#else
         saved_dir_pos = dent->d_off;
+#endif
     }
 
     /* restore (last) saved position */

--
2.32.0 (Apple Git-132)

#10 lore raw

Re: [PATCH v4 04/11] 9p: darwin: Handle struct dirent differences

Quoted 86 lines · On Montag, 7. Februar 2022 14:52:58 CET Will Cohen wrote:

On Mon, Feb 7, 2022 at 4:53 AM Fabian Franz <[email protected]>

Quoted 10 lines · wrote:

Comments inline:

hw/9pfs/9p-local.c +0 -0
Quoted 7 lines

index 1a5e3eed73..7137a28109 100644

hw/9pfs/9p-local.c +0 -0
@@ -559,6 +559,15 @@ static struct dirent *local_readdir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
Quoted 8 lines

hw/9pfs/9p-local.c static struct dirent *local_readdir(FsContext *ctx, · line 566

Quoted 6 lines
hw/9pfs/9p-local.c +3 -0
 again:
     entry = readdir(fs->dir.stream);
 
+#ifdef CONFIG_DARWIN
+    int td;
+    td = telldir(fs->dir.stream);

Maybe call this „off“?

Yes, off is better. Will adjust for v5.

Quoted 7 lines
Quoted 7 lines
hw/9pfs/9p-local.c +6 -0
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
 
Quoted 7 lines
Quoted 5 lines
     if (!entry) {
     
         return NULL;
     
     }

This needs to be before the #ifdef!

Good catch, will adjust for v5. I moved it around twice and forgot to put it in the right place.

Quoted 44 lines
Quoted 44 lines
hw/9pfs/9p-proxy.c +15 -1
index b1664080d8..8b4b5cf7dc 100644
@@ -706,7 +706,21 @@ static off_t proxy_telldir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 static struct dirent *proxy_readdir(FsContext *ctx, V9fsFidOpenState
 *fs)
 {
 
-    return readdir(fs->dir.stream);
+    struct dirent *entry;
+    entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    if (!entry) {
+        return NULL;
+    }
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
+    return entry;

}

static void proxy_seekdir(FsContext *ctx, V9fsFidOpenState *fs, off_t

off)

hw/9pfs/9p-synth.c +0 -0
index 4a4a776d06..e264a03eef 100644
@@ -222,7 +222,11 @@ static void synth_direntry(V9fsSynthNode *node,
 
 {
 
     strcpy(entry->d_name, node->name);
     entry->d_ino = node->attr->inode;
 
Quoted 7 lines
Quoted 7 lines

hw/9pfs/9p-synth.c static void synth_direntry(V9fsSynthNode *node,

Quoted 7 lines
hw/9pfs/9p-synth.c +4 -0
+#ifdef CONFIG_DARWIN
+    entry->d_seekoff = off + 1;
+#else
 
     entry->d_off = off + 1;
 
+#endif

See below ...

Quoted 37 lines
Quoted 28 lines
Quoted 28 lines

}

static struct dirent *synth_get_dentry(V9fsSynthNode *dir,

hw/9pfs/9p-util.h +12 -0
index 546f46dc7d..accbec9987 100644
@@ -79,3 +79,20 @@ ssize_t fremovexattrat_nofollow(int dirfd, const char
 *filename,
 
                                 const char *name);
 
 #endif
 
+
+
+/**
+ * Darwin has d_seekoff, which appears to function similarly to d_off.
+ * However, it does not appear to be supported on all file systems,
+ * so ensure it is manually injected earlier and call here when
+ * needed.
+ */
+
+inline off_t qemu_dirent_off(struct dirent *dent)
+{
+#ifdef CONFIG_DARWIN
Quoted 8 lines

hw/9pfs/9p-util.h

Quoted 5 lines
hw/9pfs/9p-util.h +5 -0
+    return dent->d_seekoff;
+#else
+    return dent->d_off;
+#endif
+}

Are we sure we want a helper for two times the same ifdef? Deferring to maintainers here however.

Quoted 4 lines

Either way works for me too -- my current inclination is to leave it this way (as originally suggested by the maintainers), if for no other reason than that it allows the one comment to be referenced in the case of both uses.

As requested by me in this v4, this will be 3 times in v5. So I assume that qualifies the dedicated helper function. :)

As an alternative you could make the helper function a macro instead, then you could use it in 9p-synth.c as well, which would make it 4 times. But I don't mind about that one.

Best regards,
Christian Schoenebeck

Quoted 113 lines
Quoted 24 lines
hw/9pfs/9p.c +0 -0
Quoted 23 lines

index 1563d7b7c6..cf694da354 100644

hw/9pfs/9p.c +1 -0
@@ -27,6 +27,7 @@
 
 #include "virtio-9p.h"
 #include "fsdev/qemu-fsdev.h"
 #include "9p-xattr.h"
 
+#include "9p-util.h"
 

#include "coth.h"
#include "trace.h"
#include "migration/blocker.h"

hw/9pfs/9p.c +0 -0
@@ -2281,7 +2282,11 @@ static int coroutine_fn
 v9fs_do_readdir_with_stat(V9fsPDU *pdu,
 
         count += len;
         v9fs_stat_free(&v9stat);
         v9fs_path_free(&path);
 
Quoted 9 lines

hw/9pfs/9p.c static int coroutine_fn · line 2292

Quoted 6 lines
hw/9pfs/9p.c +5 -1
-        saved_dir_pos = dent->d_off;
+        saved_dir_pos = qemu_dirent_off(dent);
+        if (saved_dir_pos < 0) {
+            err = saved_dir_pos;
+            break;
+        }

Do we still need this error-handling? I had removed it in my interdiff patch.

That's correct, it in fact can be removed. d_seekoff yields a __uint64_t (
https://developer.apple.com/documentation/kernel/direntry/1415494-d_seekoff?
language=objc). Will adjust for v5.

Quoted 24 lines
Quoted 24 lines
     }
     
     v9fs_readdir_unlock(&fidp->fs.dir);
hw/9pfs/9p.c +1 -0
@@ -2420,6 +2425,7 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp,
 
     V9fsString name;
     int len, err = 0;
     int32_t count = 0;
 
+    off_t off;
     struct dirent *dent;
     struct stat *st;
     struct V9fsDirEnt *entries = NULL;
hw/9pfs/9p.c +0 -0
@@ -2480,12 +2486,17 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp,
 
             qid.version = 0;
 
         }
 
Quoted 7 lines

hw/9pfs/9p.c static int coroutine_fn v9fs_do_readdir(V9fsPDU · line 2496

Quoted 5 lines
hw/9pfs/9p.c +5 -0
+        off = qemu_dirent_off(dent);
+        if (off < 0) {
+            err = off;
+            break;
+        }

Same here - if this can never fail, why add the error handling?

See above.

Quoted 41 lines
Quoted 41 lines
hw/9pfs/9p.c +1 -1
         v9fs_string_init(&name);
         v9fs_string_sprintf(&name, "%s", dent->d_name);
 
         /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
         len = pdu_marshal(pdu, 11 + count, "Qqbs",
 
-                          &qid, dent->d_off,
+                          &qid, off,
 
                           dent->d_type, &name);
 
         v9fs_string_free(&name);
hw/9pfs/codir.c +7 -1
index 032cce04c4..fac6759a64 100644
@@ -167,7 +167,14 @@ static int do_readdir_many(V9fsPDU *pdu,
 V9fsFidState *fidp,
 
         }
 
         size += len;
 
+        /* This conditional statement is identical in
+         * function to qemu_dirent_off, described in 9p-util.h,
+         * since that header cannot be included here. */
+#ifdef CONFIG_DARWIN
+        saved_dir_pos = dent->d_seekoff;
+#else
 
         saved_dir_pos = dent->d_off;
 
+#endif
 
     }
 
     /* restore (last) saved position */
 
--

2.32.0 (Apple Git-132)

#11 lore raw

Re: [PATCH v4 04/11] 9p: darwin: Handle struct dirent differences

In reply to Will Cohen
Quoted 48 lines · On Sonntag, 6. Februar 2022 21:07:12 CET Will Cohen wrote:

From: Keno Fischer <[email protected]>

On darwin d_seekoff exists, but is optional and does not seem to be commonly used by file systems. Use telldir instead to obtain the seek offset and inject it into d_seekoff, and create a qemu_dirent_off helper to call it appropriately when appropriate.

[Michael Roitzsch: - Rebase for NixOS]

[Will Cohen: - Adjust to pass testing
- Ensure that d_seekoff is filled using telldir on darwin, and create qemu_dirent_off helper to decide which to access]
[Fabian Franz: - Add telldir error handling for darwin] [Will Cohen: - Ensure that telldir error handling uses
signed int]

---
 hw/9pfs/9p-local.c |  9 +++++++++
 hw/9pfs/9p-proxy.c | 16 +++++++++++++++-
 hw/9pfs/9p-synth.c |  4 ++++
 hw/9pfs/9p-util.h  | 17 +++++++++++++++++
 hw/9pfs/9p.c       | 15 +++++++++++++--
 hw/9pfs/codir.c    |  7 +++++++
 6 files changed, 65 insertions(+), 3 deletions(-)
hw/9pfs/9p-local.c +9 -0
index 1a5e3eed73..7137a28109 100644
@@ -559,6 +559,15 @@ static struct dirent *local_readdir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 again:
     entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
     if (!entry) {
         return NULL;
Quoted 1 line
     }

'entry' may be NULL, so the 'if (!entry) {' check should be before the Darwin specific code to avoid a crash on macOS.

Quoted 51 lines
hw/9pfs/9p-proxy.c +15 -1
index b1664080d8..8b4b5cf7dc 100644
@@ -706,7 +706,21 @@ static off_t proxy_telldir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 static struct dirent *proxy_readdir(FsContext *ctx, V9fsFidOpenState *fs)
 {
-    return readdir(fs->dir.stream);
+    struct dirent *entry;
+    entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    if (!entry) {
+        return NULL;
+    }
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
+    return entry;
 }
 

static void proxy_seekdir(FsContext *ctx, V9fsFidOpenState *fs, off_t off)

hw/9pfs/9p-synth.c +4 -0
index 4a4a776d06..e264a03eef 100644
@@ -222,7 +222,11 @@ static void synth_direntry(V9fsSynthNode *node,
 {
     strcpy(entry->d_name, node->name);
     entry->d_ino = node->attr->inode;
+#ifdef CONFIG_DARWIN
+    entry->d_seekoff = off + 1;
+#else
     entry->d_off = off + 1;
+#endif
 }
 
 static struct dirent *synth_get_dentry(V9fsSynthNode *dir,
hw/9pfs/9p-util.h +0 -0
index 546f46dc7d..accbec9987 100644
@@ -79,3 +79,20 @@ ssize_t fremovexattrat_nofollow(int dirfd, const char
 *filename, const char *name);
 

hw/9pfs/9p-util.h ssize_t fremovexattrat_nofollow(int dirfd, const char · line 90

Quoted 10 lines
hw/9pfs/9p-util.h +9 -0
 #endif
+
+
+/**
+ * Darwin has d_seekoff, which appears to function similarly to d_off.
+ * However, it does not appear to be supported on all file systems,
+ * so ensure it is manually injected earlier and call here when
+ * needed.
+ */
+

Nitpicking: no blank line here please.

Quoted 21 lines
hw/9pfs/9p-util.h +8 -0
+inline off_t qemu_dirent_off(struct dirent *dent)
+{
+#ifdef CONFIG_DARWIN
+    return dent->d_seekoff;
+#else
+    return dent->d_off;
+#endif
+}
hw/9pfs/9p.c +1 -0
index 1563d7b7c6..cf694da354 100644
@@ -27,6 +27,7 @@
 #include "virtio-9p.h"
 #include "fsdev/qemu-fsdev.h"
 #include "9p-xattr.h"
+#include "9p-util.h"
 #include "coth.h"
 #include "trace.h"
 #include "migration/blocker.h"

hw/9pfs/9p.c static int coroutine_fn · line 2290

Quoted 10 lines
hw/9pfs/9p.c +5 -1
@@ -2281,7 +2282,11 @@ static int coroutine_fn
 v9fs_do_readdir_with_stat(V9fsPDU *pdu, count += len;
         v9fs_stat_free(&v9stat);
         v9fs_path_free(&path);
-        saved_dir_pos = dent->d_off;
+        saved_dir_pos = qemu_dirent_off(dent);
+        if (saved_dir_pos < 0) {
+            err = saved_dir_pos;
+            break;
+        }
     }

That check is no longer needed here, is it?

Quoted 11 lines
hw/9pfs/9p.c +1 -0
     v9fs_readdir_unlock(&fidp->fs.dir);
@@ -2420,6 +2425,7 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU *pdu,
 V9fsFidState *fidp, V9fsString name;
     int len, err = 0;
     int32_t count = 0;
+    off_t off;
     struct dirent *dent;
     struct stat *st;
     struct V9fsDirEnt *entries = NULL;

hw/9pfs/9p.c static int coroutine_fn v9fs_do_readdir(V9fsPDU · line 2493

Quoted 8 lines
hw/9pfs/9p.c +5 -0
@@ -2480,12 +2486,17 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp, qid.version = 0;
         }
 
+        off = qemu_dirent_off(dent);
+        if (off < 0) {
+            err = off;
+            break;
+        }

Likewise: is this check still needed?

Quoted 16 lines
hw/9pfs/9p.c +1 -1
         v9fs_string_init(&name);
         v9fs_string_sprintf(&name, "%s", dent->d_name);
 
         /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
         len = pdu_marshal(pdu, 11 + count, "Qqbs",
-                          &qid, dent->d_off,
+                          &qid, off,
                           dent->d_type, &name);
 
         v9fs_string_free(&name);
hw/9pfs/codir.c +0 -0
index 032cce04c4..fac6759a64 100644
@@ -167,7 +167,14 @@ static int do_readdir_many(V9fsPDU *pdu, V9fsFidState
 *fidp, }

hw/9pfs/codir.c static int do_readdir_many(V9fsPDU *pdu, V9fsFidState · line 177

Quoted 10 lines
hw/9pfs/codir.c +7 -0
 
         size += len;
+        /* This conditional statement is identical in
+         * function to qemu_dirent_off, described in 9p-util.h,
+         * since that header cannot be included here. */
+#ifdef CONFIG_DARWIN
+        saved_dir_pos = dent->d_seekoff;
+#else
         saved_dir_pos = dent->d_off;
+#endif

Why can't the header not be included here? Obvious preference would be to use qemu_dirent_off() here as well, to have control at one central code location.

Quoted 3 lines
hw/9pfs/codir.c +0 -0
     }
 
     /* restore (last) saved position */
author
#12 lore raw

Re: [PATCH v4 04/11] 9p: darwin: Handle struct dirent differences

Quoted 46 lines · Christian Schoenebeck
Quoted 45 lines · On Sonntag, 6. Februar 2022 21:07:12 CET Will Cohen wrote:

From: Keno Fischer <[email protected]>

On darwin d_seekoff exists, but is optional and does not seem to be commonly used by file systems. Use telldir instead to obtain the seek offset and inject it into d_seekoff, and create a qemu_dirent_off helper to call it appropriately when appropriate.

[Michael Roitzsch: - Rebase for NixOS]

[Will Cohen: - Adjust to pass testing
- Ensure that d_seekoff is filled using telldir on darwin, and create qemu_dirent_off helper to decide which to access]
[Fabian Franz: - Add telldir error handling for darwin] [Will Cohen: - Ensure that telldir error handling uses
signed int]

---
 hw/9pfs/9p-local.c |  9 +++++++++
 hw/9pfs/9p-proxy.c | 16 +++++++++++++++-
 hw/9pfs/9p-synth.c |  4 ++++
 hw/9pfs/9p-util.h  | 17 +++++++++++++++++
 hw/9pfs/9p.c       | 15 +++++++++++++--
 hw/9pfs/codir.c    |  7 +++++++
 6 files changed, 65 insertions(+), 3 deletions(-)
hw/9pfs/9p-local.c +8 -0
index 1a5e3eed73..7137a28109 100644
@@ -559,6 +559,15 @@ static struct dirent *local_readdir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 again:
     entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
Quoted 9 lines
Quoted 4 lines
hw/9pfs/9p-local.c +1 -0
+#endif
     if (!entry) {
         return NULL;
     }

'entry' may be NULL, so the 'if (!entry) {' check should be before the Darwin
specific code to avoid a crash on macOS.

Adjusting for v5.

Quoted 54 lines

hw/9pfs/9p-proxy.c static off_t proxy_telldir(FsContext *ctx, · line 708

Quoted 9 lines
hw/9pfs/9p-proxy.c +0 -0
index b1664080d8..8b4b5cf7dc 100644
@@ -706,7 +706,21 @@ static off_t proxy_telldir(FsContext *ctx,
 V9fsFidOpenState *fs)
 
 static struct dirent *proxy_readdir(FsContext *ctx, V9fsFidOpenState

*fs)

Quoted 19 lines
hw/9pfs/9p-proxy.c +15 -1
 {
-    return readdir(fs->dir.stream);
+    struct dirent *entry;
+    entry = readdir(fs->dir.stream);
+#ifdef CONFIG_DARWIN
+    if (!entry) {
+        return NULL;
+    }
+    int td;
+    td = telldir(fs->dir.stream);
+    /* If telldir fails, fail the entire readdir call */
+    if (td < 0) {
+        return NULL;
+    }
+    entry->d_seekoff = td;
+#endif
+    return entry;
 }
 
Quoted 1 line

static void proxy_seekdir(FsContext *ctx, V9fsFidOpenState *fs, off_t

off)

Quoted 23 lines
hw/9pfs/9p-synth.c +4 -0
index 4a4a776d06..e264a03eef 100644
@@ -222,7 +222,11 @@ static void synth_direntry(V9fsSynthNode *node,
 {
     strcpy(entry->d_name, node->name);
     entry->d_ino = node->attr->inode;
+#ifdef CONFIG_DARWIN
+    entry->d_seekoff = off + 1;
+#else
     entry->d_off = off + 1;
+#endif
 }
 
 static struct dirent *synth_get_dentry(V9fsSynthNode *dir,
hw/9pfs/9p-util.h +0 -0
index 546f46dc7d..accbec9987 100644
@@ -79,3 +79,20 @@ ssize_t fremovexattrat_nofollow(int dirfd, const char
 *filename, const char *name);
 
Quoted 13 lines

hw/9pfs/9p-util.h ssize_t fremovexattrat_nofollow(int dirfd, const char *filename, · line 90

Quoted 10 lines
hw/9pfs/9p-util.h +9 -0
 #endif
+
+
+/**
+ * Darwin has d_seekoff, which appears to function similarly to d_off.
+ * However, it does not appear to be supported on all file systems,
+ * so ensure it is manually injected earlier and call here when
+ * needed.
+ */
+

Nitpicking: no blank line here please.

Adjusting for v5.

Quoted 26 lines
Quoted 26 lines
hw/9pfs/9p-util.h +8 -0
+inline off_t qemu_dirent_off(struct dirent *dent)
+{
+#ifdef CONFIG_DARWIN
+    return dent->d_seekoff;
+#else
+    return dent->d_off;
+#endif
+}
hw/9pfs/9p.c +1 -1
index 1563d7b7c6..cf694da354 100644
@@ -27,6 +27,7 @@
 #include "virtio-9p.h"
 #include "fsdev/qemu-fsdev.h"
 #include "9p-xattr.h"
+#include "9p-util.h"
 #include "coth.h"
 #include "trace.h"
 #include "migration/blocker.h"
@@ -2281,7 +2282,11 @@ static int coroutine_fn
 v9fs_do_readdir_with_stat(V9fsPDU *pdu, count += len;
         v9fs_stat_free(&v9stat);
         v9fs_path_free(&path);
-        saved_dir_pos = dent->d_off;
Quoted 9 lines

hw/9pfs/9p.c static int coroutine_fn v9fs_do_readdir_with_stat(V9fsPDU *pdu, · line 2290

Quoted 6 lines
hw/9pfs/9p.c +5 -0
+        saved_dir_pos = qemu_dirent_off(dent);
+        if (saved_dir_pos < 0) {
+            err = saved_dir_pos;
+            break;
+        }
     }

That check is no longer needed here, is it?

Correct! Adjusting for v5.

Quoted 16 lines

hw/9pfs/9p.c static int coroutine_fn v9fs_do_readdir(V9fsPDU

Quoted 4 lines
hw/9pfs/9p.c +0 -0
     v9fs_readdir_unlock(&fidp->fs.dir);
@@ -2420,6 +2425,7 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU

*pdu,

Quoted 11 lines
hw/9pfs/9p.c +1 -0
 V9fsFidState *fidp, V9fsString name;
     int len, err = 0;
     int32_t count = 0;
+    off_t off;
     struct dirent *dent;
     struct stat *st;
     struct V9fsDirEnt *entries = NULL;
@@ -2480,12 +2486,17 @@ static int coroutine_fn v9fs_do_readdir(V9fsPDU
 *pdu, V9fsFidState *fidp, qid.version = 0;
         }
 
Quoted 8 lines

hw/9pfs/9p.c static int coroutine_fn v9fs_do_readdir(V9fsPDU *pdu, V9fsFidState *fidp, · line 2493

Quoted 5 lines
hw/9pfs/9p.c +5 -0
+        off = qemu_dirent_off(dent);
+        if (off < 0) {
+            err = off;
+            break;
+        }

Likewise: is this check still needed?

As above, removing for v5.

Quoted 19 lines

hw/9pfs/codir.c static int do_readdir_many(V9fsPDU *pdu,

Quoted 16 lines
hw/9pfs/9p.c +1 -1
         v9fs_string_init(&name);
         v9fs_string_sprintf(&name, "%s", dent->d_name);
 
         /* 11 = 7 + 4 (7 = start offset, 4 = space for storing count) */
         len = pdu_marshal(pdu, 11 + count, "Qqbs",
-                          &qid, dent->d_off,
+                          &qid, off,
                           dent->d_type, &name);
 
         v9fs_string_free(&name);
hw/9pfs/codir.c +0 -0
index 032cce04c4..fac6759a64 100644
@@ -167,7 +167,14 @@ static int do_readdir_many(V9fsPDU *pdu,

V9fsFidState

Quoted 2 lines
hw/9pfs/codir.c +0 -0
 *fidp, }
 
Quoted 15 lines

hw/9pfs/codir.c static int do_readdir_many(V9fsPDU *pdu, V9fsFidState *fidp, · line 177

Quoted 9 lines
hw/9pfs/codir.c +7 -0
         size += len;
+        /* This conditional statement is identical in
+         * function to qemu_dirent_off, described in 9p-util.h,
+         * since that header cannot be included here. */
+#ifdef CONFIG_DARWIN
+        saved_dir_pos = dent->d_seekoff;
+#else
         saved_dir_pos = dent->d_off;
+#endif

Why can't the header not be included here? Obvious preference would be to use
qemu_dirent_off() here as well, to have control at one central code location.

Only my own imprecision in organizing the includes correctly. After including "9p-xattr.h" as well, this works. Will adjust for v5.

Quoted 7 lines
Quoted 4 lines
hw/9pfs/codir.c +0 -0
     }
 
     /* restore (last) saved position */
author
#13 lore raw

PATCH v4 5/11 9p: darwin: Ignore O_{NOATIME, DIRECT}

In reply to Will Cohen

From: Keno Fischer <[email protected]>

Darwin doesn't have either of these flags. Darwin does have F_NOCACHE, which is similar to O_DIRECT, but has different enough semantics that other projects don't generally map them automatically. In any case, we don't support O_DIRECT on Linux at the moment either.

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Adjust coding style]
---
 hw/9pfs/9p-util.h |  2 ++
 hw/9pfs/9p.c      | 13 ++++++++++++-
 2 files changed, 14 insertions(+), 1 deletion(-)
hw/9pfs/9p-util.h +2 -0
index accbec9987..3b9c485414 100644
@@ -41,6 +41,7 @@ again:
     fd = openat(dirfd, name, flags | O_NOFOLLOW | O_NOCTTY | O_NONBLOCK,
                 mode);
     if (fd == -1) {
+#ifndef CONFIG_DARWIN
         if (errno == EPERM && (flags & O_NOATIME)) {
             /*
              * The client passed O_NOATIME but we lack permissions to honor it.
@@ -53,6 +54,7 @@ again:
             flags &= ~O_NOATIME;
             goto again;
         }
+#endif
         return -1;
     }
 
hw/9pfs/9p.c +12 -1
index cf694da354..f3b00d20a2 100644
@@ -138,11 +138,20 @@ static int dotl_to_open_flags(int flags)
         { P9_DOTL_NONBLOCK, O_NONBLOCK } ,
         { P9_DOTL_DSYNC, O_DSYNC },
         { P9_DOTL_FASYNC, FASYNC },
+#ifndef CONFIG_DARWIN
+        { P9_DOTL_NOATIME, O_NOATIME },
+        /*
+         *  On Darwin, we could map to F_NOCACHE, which is
+         *  similar, but doesn't quite have the same
+         *  semantics. However, we don't support O_DIRECT
+         *  even on linux at the moment, so we just ignore
+         *  it here.
+         */
         { P9_DOTL_DIRECT, O_DIRECT },
+#endif
         { P9_DOTL_LARGEFILE, O_LARGEFILE },
         { P9_DOTL_DIRECTORY, O_DIRECTORY },
         { P9_DOTL_NOFOLLOW, O_NOFOLLOW },
-        { P9_DOTL_NOATIME, O_NOATIME },
         { P9_DOTL_SYNC, O_SYNC },
     };
 
@@ -171,10 +180,12 @@ static int get_dotl_openflags(V9fsState *s, int oflags)
      */
     flags = dotl_to_open_flags(oflags);
     flags &= ~(O_NOCTTY | O_ASYNC | O_CREAT);
+#ifndef CONFIG_DARWIN
     /*
      * Ignore direct disk access hint until the server supports it.
      */
     flags &= ~O_DIRECT;
+#endif
     return flags;
 }
 
Signature · 1 line
2.32.0 (Apple Git-132)
author
#14 lore raw

PATCH v4 6/11 9p: darwin: Move XATTR_SIZE_MAX->P9_XATTR_SIZE_MAX

In reply to Will Cohen

From: Keno Fischer <[email protected]>

Because XATTR_SIZE_MAX is not defined on Darwin, create a cross-platform P9_XATTR_SIZE_MAX instead.

[Will Cohen: - Adjust coding style
- Lower XATTR_SIZE_MAX to 64k
- Add explanatory context related to XATTR_SIZE_MAX]
[Fabian Franz: - Move XATTR_SIZE_MAX reference from 9p.c to

                 P9_XATTR_SIZE_MAX in 9p.h]
---
 hw/9pfs/9p.c |  2 +-
 hw/9pfs/9p.h | 11 +++++++++++
 2 files changed, 12 insertions(+), 1 deletion(-)
hw/9pfs/9p.c +1 -1
index f3b00d20a2..440bf5c9e5 100644
@@ -3957,7 +3957,7 @@ static void coroutine_fn v9fs_xattrcreate(void *opaque)
         rflags |= XATTR_REPLACE;
     }
 
-    if (size > XATTR_SIZE_MAX) {
+    if (size > P9_XATTR_SIZE_MAX) {
         err = -E2BIG;
         goto out_nofid;
     }
hw/9pfs/9p.h +11 -0
index 1567b67841..6a1856b4dc 100644
@@ -479,4 +479,15 @@ struct V9fsTransport {
     void        (*push_and_notify)(V9fsPDU *pdu);
 };
 
+/*
+ * Darwin doesn't seem to define a maximum xattr size in its user
+ * space header, so manually configure it across platforms as 64k.
+ *
+ * Having no limit at all can lead to QEMU crashing during large g_malloc()
+ * calls. Because QEMU does not currently support macOS guests, the below
+ * preliminary solution only works due to its being a reflection of the limit of
+ * Linux guests.
+ */
+#define P9_XATTR_SIZE_MAX 65536
+
 #endif
Signature · 1 line
2.32.0 (Apple Git-132)
author
#15 lore raw

PATCH v4 7/11 9p: darwin: *xattr_nofollow implementations

In reply to Will Cohen

From: Keno Fischer <[email protected]>

This implements the darwin equivalent of the functions that were moved to 9p-util(-linux) earlier in this series in the new 9p-util-darwin file.

[Michael Roitzsch: - Rebase for NixOS]
---
 hw/9pfs/9p-util-darwin.c | 64 ++++++++++++++++++++++++++++++++++++++++
 hw/9pfs/meson.build      |  1 +
 2 files changed, 65 insertions(+)
 create mode 100644 hw/9pfs/9p-util-darwin.c
hw/9pfs/9p-util-darwin.c new file +64 -0
new file mode 100644
index 0000000000..cdb4c9e24c
@@ -0,0 +1,64 @@
+/*
+ * 9p utilities (Darwin Implementation)
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2 or later.
+ * See the COPYING file in the top-level directory.
+ */
+
+#include "qemu/osdep.h"
+#include "qemu/xattr.h"
+#include "9p-util.h"
+
+ssize_t fgetxattrat_nofollow(int dirfd, const char *filename, const char *name,
+                             void *value, size_t size)
+{
+    int ret;
+    int fd = openat_file(dirfd, filename,
+                         O_RDONLY | O_PATH_9P_UTIL | O_NOFOLLOW, 0);
+    if (fd == -1) {
+        return -1;
+    }
+    ret = fgetxattr(fd, name, value, size, 0, 0);
+    close_preserve_errno(fd);
+    return ret;
+}
+
+ssize_t flistxattrat_nofollow(int dirfd, const char *filename,
+                              char *list, size_t size)
+{
+    int ret;
+    int fd = openat_file(dirfd, filename,
+                         O_RDONLY | O_PATH_9P_UTIL | O_NOFOLLOW, 0);
+    if (fd == -1) {
+        return -1;
+    }
+    ret = flistxattr(fd, list, size, 0);
+    close_preserve_errno(fd);
+    return ret;
+}
+
+ssize_t fremovexattrat_nofollow(int dirfd, const char *filename,
+                                const char *name)
+{
+    int ret;
+    int fd = openat_file(dirfd, filename, O_PATH_9P_UTIL | O_NOFOLLOW, 0);
+    if (fd == -1) {
+        return -1;
+    }
+    ret = fremovexattr(fd, name, 0);
+    close_preserve_errno(fd);
+    return ret;
+}
+
+int fsetxattrat_nofollow(int dirfd, const char *filename, const char *name,
+                         void *value, size_t size, int flags)
+{
+    int ret;
+    int fd = openat_file(dirfd, filename, O_PATH_9P_UTIL | O_NOFOLLOW, 0);
+    if (fd == -1) {
+        return -1;
+    }
+    ret = fsetxattr(fd, name, value, size, 0, flags);
+    close_preserve_errno(fd);
+    return ret;
+}
hw/9pfs/meson.build +1 -0
index 1b28e70040..12443b6ad5 100644
@@ -14,6 +14,7 @@ fs_ss.add(files(
   'coxattr.c',
 ))
 fs_ss.add(when: 'CONFIG_LINUX', if_true: files('9p-util-linux.c'))
+fs_ss.add(when: 'CONFIG_DARWIN', if_true: files('9p-util-darwin.c'))
 fs_ss.add(when: 'CONFIG_XEN', if_true: files('xen-9p-backend.c'))
 softmmu_ss.add_all(when: 'CONFIG_FSDEV_9P', if_true: fs_ss)
 
Signature · 1 line
2.32.0 (Apple Git-132)
author
#16 lore raw

PATCH v4 8/11 9p: darwin: Compatibility for f/l*xattr

In reply to Will Cohen

From: Keno Fischer <[email protected]>

On darwin fgetxattr takes two extra optional arguments, and the l* variants are not defined (in favor of an extra flag to the regular variants.

[Michael Roitzsch: - Rebase for NixOS]
---
 hw/9pfs/9p-local.c | 12 ++++++++----
 hw/9pfs/9p-util.h  | 17 +++++++++++++++++
 2 files changed, 25 insertions(+), 4 deletions(-)
hw/9pfs/9p-local.c +8 -4
index 7137a28109..ea3ded5fca 100644
@@ -790,16 +790,20 @@ static int local_fstat(FsContext *fs_ctx, int fid_type,
         mode_t tmp_mode;
         dev_t tmp_dev;
 
-        if (fgetxattr(fd, "user.virtfs.uid", &tmp_uid, sizeof(uid_t)) > 0) {
+        if (qemu_fgetxattr(fd, "user.virtfs.uid",
+                           &tmp_uid, sizeof(uid_t)) > 0) {
             stbuf->st_uid = le32_to_cpu(tmp_uid);
         }
-        if (fgetxattr(fd, "user.virtfs.gid", &tmp_gid, sizeof(gid_t)) > 0) {
+        if (qemu_fgetxattr(fd, "user.virtfs.gid",
+                           &tmp_gid, sizeof(gid_t)) > 0) {
             stbuf->st_gid = le32_to_cpu(tmp_gid);
         }
-        if (fgetxattr(fd, "user.virtfs.mode", &tmp_mode, sizeof(mode_t)) > 0) {
+        if (qemu_fgetxattr(fd, "user.virtfs.mode",
+                           &tmp_mode, sizeof(mode_t)) > 0) {
             stbuf->st_mode = le32_to_cpu(tmp_mode);
         }
-        if (fgetxattr(fd, "user.virtfs.rdev", &tmp_dev, sizeof(dev_t)) > 0) {
+        if (qemu_fgetxattr(fd, "user.virtfs.rdev",
+                           &tmp_dev, sizeof(dev_t)) > 0) {
             stbuf->st_rdev = le64_to_cpu(tmp_dev);
         }
     } else if (fs_ctx->export_flags & V9FS_SM_MAPPED_FILE) {
hw/9pfs/9p-util.h +17 -0
index 3b9c485414..8e610ad224 100644
@@ -19,6 +19,23 @@
 #define O_PATH_9P_UTIL 0
 #endif
 
+#ifdef CONFIG_DARWIN
+#define qemu_fgetxattr(...) fgetxattr(__VA_ARGS__, 0, 0)
+#define qemu_lgetxattr(...) getxattr(__VA_ARGS__, 0, XATTR_NOFOLLOW)
+#define qemu_llistxattr(...) listxattr(__VA_ARGS__, XATTR_NOFOLLOW)
+#define qemu_lremovexattr(...) removexattr(__VA_ARGS__, XATTR_NOFOLLOW)
+static inline int qemu_lsetxattr(const char *path, const char *name,
+                                 const void *value, size_t size, int flags) {
+    return setxattr(path, name, value, size, 0, flags | XATTR_NOFOLLOW);
+}
+#else
+#define qemu_fgetxattr fgetxattr
+#define qemu_lgetxattr lgetxattr
+#define qemu_llistxattr llistxattr
+#define qemu_lremovexattr lremovexattr
+#define qemu_lsetxattr lsetxattr
+#endif
+
 static inline void close_preserve_errno(int fd)
 {
     int serrno = errno;
Signature · 1 line
2.32.0 (Apple Git-132)
author
#17 lore raw

PATCH v4 9/11 9p: darwin: Implement compatibility for mknodat

In reply to Will Cohen

From: Keno Fischer <[email protected]>

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly]

---
 hw/9pfs/9p-local.c       |  5 +++--
 hw/9pfs/9p-util-darwin.c | 27 +++++++++++++++++++++++++++
 hw/9pfs/9p-util-linux.c  |  5 +++++
 hw/9pfs/9p-util.h        |  2 ++
 4 files changed, 37 insertions(+), 2 deletions(-)
hw/9pfs/9p-local.c +3 -2
index ea3ded5fca..0569a8b4de 100644
@@ -682,7 +682,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath *dir_path,
 
     if (fs_ctx->export_flags & V9FS_SM_MAPPED ||
         fs_ctx->export_flags & V9FS_SM_MAPPED_FILE) {
-        err = mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
+        err = qemu_mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
         if (err == -1) {
             goto out;
         }
@@ -697,7 +697,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath *dir_path,
         }
     } else if (fs_ctx->export_flags & V9FS_SM_PASSTHROUGH ||
                fs_ctx->export_flags & V9FS_SM_NONE) {
-        err = mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
+        err = qemu_mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
         if (err == -1) {
             goto out;
         }
@@ -710,6 +710,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath *dir_path,
 
 err_end:
     unlinkat_preserve_errno(dirfd, name, 0);
+
 out:
     close_preserve_errno(dirfd);
     return err;
hw/9pfs/9p-util-darwin.c +27 -0
index cdb4c9e24c..128b6d87e8 100644
@@ -5,6 +5,7 @@
  * See the COPYING file in the top-level directory.
  */
 
+#include <os/availability.h>
 #include "qemu/osdep.h"
 #include "qemu/xattr.h"
 #include "9p-util.h"
@@ -62,3 +63,29 @@ int fsetxattrat_nofollow(int dirfd, const char *filename, const char *name,
     close_preserve_errno(fd);
     return ret;
 }
+
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed.
+ *
+ * Radar filed with Apple for implementing mknodat:
+ * rdar://FB9862426 (https://openradar.appspot.com/FB9862426)
+ */
+
+int pthread_fchdir_np(int fd) API_AVAILABLE(macosx(10.12));
+
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    int preserved_errno, err;
+    if (pthread_fchdir_np(dirfd) < 0) {
+        return -1;
+    }
+    err = mknod(filename, mode, dev);
+    preserved_errno = errno;
+    /* Stop using the thread-local cwd */
+    pthread_fchdir_np(-1);
+    if (err < 0) {
+        errno = preserved_errno;
+    }
+    return err;
+}
hw/9pfs/9p-util-linux.c +5 -0
index 398614a5d0..66e0ab1865 100644
@@ -62,3 +62,8 @@ int fsetxattrat_nofollow(int dirfd, const char *filename, const char *name,
     g_free(proc_path);
     return ret;
 }
+
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    return mknodat(dirfd, filename, mode, dev);
+}
hw/9pfs/9p-util.h +2 -0
index 8e610ad224..f6fed963bf 100644
@@ -97,6 +97,8 @@ ssize_t flistxattrat_nofollow(int dirfd, const char *filename,
 ssize_t fremovexattrat_nofollow(int dirfd, const char *filename,
                                 const char *name);
 
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev);
+
 #endif
 
 
Signature · 1 line
2.32.0 (Apple Git-132)
#18 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 33 lines · Will Cohen

From: Keno Fischer <[email protected]>

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly]

---
  hw/9pfs/9p-local.c       |  5 +++--
  hw/9pfs/9p-util-darwin.c | 27 +++++++++++++++++++++++++++
  hw/9pfs/9p-util-linux.c  |  5 +++++
  hw/9pfs/9p-util.h        |  2 ++
  4 files changed, 37 insertions(+), 2 deletions(-)
hw/9pfs/9p-util.h +0 -0
index 8e610ad224..f6fed963bf 100644

hw/9pfs/9p-util.h ssize_t flistxattrat_nofollow(int dirfd, const char *filename, · line 100

Quoted 4 lines
hw/9pfs/9p-util.h +1 -0
@@ -97,6 +97,8 @@ ssize_t flistxattrat_nofollow(int dirfd, const char *filename,
  ssize_t fremovexattrat_nofollow(int dirfd, const char *filename,
                                  const char *name);
 
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev);

I think this belong to "osdep.h" & os-posix.c.

author
#19 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

This patch set currently places it in 9p-util only because 9p is the only place where this issue seems to have come up so far and we were wary of editing files too far afield, but I have no attachment to its specific location!

Quoted 43 lines · Philippe Mathieu-Daudé

hw/9pfs/9p-util.h ssize_t flistxattrat_nofollow(int dirfd, const char

Quoted 33 lines · Will Cohen

From: Keno Fischer <[email protected]>

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly]

---
  hw/9pfs/9p-local.c       |  5 +++--
  hw/9pfs/9p-util-darwin.c | 27 +++++++++++++++++++++++++++
  hw/9pfs/9p-util-linux.c  |  5 +++++
  hw/9pfs/9p-util.h        |  2 ++
  4 files changed, 37 insertions(+), 2 deletions(-)
hw/9pfs/9p-util.h +0 -0
index 8e610ad224..f6fed963bf 100644
@@ -97,6 +97,8 @@ ssize_t flistxattrat_nofollow(int dirfd, const char

*filename,

hw/9pfs/9p-util.h ssize_t flistxattrat_nofollow(int dirfd, const char *filename,

Quoted 4 lines
hw/9pfs/9p-util.h +1 -0
  ssize_t fremovexattrat_nofollow(int dirfd, const char *filename,
                                  const char *name);
 
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t

dev);

I think this belong to "osdep.h" & os-posix.c.

#20 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 5 lines · On Sun, 6 Feb 2022 20:10:23 -0500 Will Cohen <[email protected]> wrote:

This patch set currently places it in 9p-util only because 9p is the only place where this issue seems to have come up so far and we were wary of editing files too far afield, but I have no attachment to its specific location!

Inline comments are preferred on qemu-devel. Please don't top post ! This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's some other user in the QEMU code base, it is acceptable to leave it under 9pfs.

Quoted 46 lines
Quoted 43 lines · Philippe Mathieu-Daudé

hw/9pfs/9p-util.h ssize_t flistxattrat_nofollow(int dirfd, const char

Quoted 33 lines · Will Cohen

From: Keno Fischer <[email protected]>

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly]

---
  hw/9pfs/9p-local.c       |  5 +++--
  hw/9pfs/9p-util-darwin.c | 27 +++++++++++++++++++++++++++
  hw/9pfs/9p-util-linux.c  |  5 +++++
  hw/9pfs/9p-util.h        |  2 ++
  4 files changed, 37 insertions(+), 2 deletions(-)
hw/9pfs/9p-util.h +0 -0
index 8e610ad224..f6fed963bf 100644
@@ -97,6 +97,8 @@ ssize_t flistxattrat_nofollow(int dirfd, const char

*filename,

hw/9pfs/9p-util.h ssize_t flistxattrat_nofollow(int dirfd, const char *filename,

Quoted 4 lines
hw/9pfs/9p-util.h +1 -0
  ssize_t fremovexattrat_nofollow(int dirfd, const char *filename,
                                  const char *name);
 
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t

dev);

I think this belong to "osdep.h" & os-posix.c.

#21 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 9 lines · Greg Kurz
Quoted 5 lines · On Sun, 6 Feb 2022 20:10:23 -0500 Will Cohen <[email protected]> wrote:

This patch set currently places it in 9p-util only because 9p is the only place where this issue seems to have come up so far and we were wary of editing files too far afield, but I have no attachment to its specific location!

Quoted 6 lines

Inline comments are preferred on qemu-devel. Please don't top post ! This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's some other user in the QEMU code base, it is acceptable to leave it under 9pfs.

virtiofsd could eventually use it.

#22 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 10 lines · On Mon, 7 Feb 2022 11:30:18 +0100 Philippe Mathieu-Daudé <[email protected]>…
Quoted 9 lines · Greg Kurz
Quoted 5 lines · On Sun, 6 Feb 2022 20:10:23 -0500 Will Cohen <[email protected]> wrote:

This patch set currently places it in 9p-util only because 9p is the only place where this issue seems to have come up so far and we were wary of editing files too far afield, but I have no attachment to its specific location!

Quoted 8 lines
Quoted 6 lines

Inline comments are preferred on qemu-devel. Please don't top post ! This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's some other user in the QEMU code base, it is acceptable to leave it under 9pfs.

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not aware of any work to support any other host OS.

Cc'ing virtio-fs people for inputs on this topic.

#23 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 20 lines · * Greg Kurz ([email protected]) wrote:
Quoted 17 lines · On Mon, 7 Feb 2022 11:30:18 +0100 Philippe Mathieu-Daudé <[email protected]>…
Quoted 15 lines · Greg Kurz
Quoted 5 lines · On Sun, 6 Feb 2022 20:10:23 -0500 Will Cohen <[email protected]> wrote:

This patch set currently places it in 9p-util only because 9p is the only place where this issue seems to have come up so far and we were wary of editing files too far afield, but I have no attachment to its specific location!

Inline comments are preferred on qemu-devel. Please don't top post ! This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's some other user in the QEMU code base, it is acceptable to leave it under 9pfs.

Quoted 7 lines
Quoted 1 line

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not aware of any work to support any other host OS.

Cc'ing virtio-fs people for inputs on this topic.

Indeeed, there's a lot of Linux specific code in the virtiofsd - I know people are interested in other platforms, but I'm not sure that's the right starting point.

Dave

Signature · 1 line
Dr. David Alan Gilbert / [email protected] / Manchester, UK
#24 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 30 lines · On Montag, 7. Februar 2022 11:57:25 CET Dr. David Alan Gilbert wrote:
Quoted 29 lines · * Greg Kurz ([email protected]) wrote:

On Mon, 7 Feb 2022 11:30:18 +0100

Quoted 15 lines · Philippe Mathieu-Daudé
Quoted 14 lines · Greg Kurz

On Sun, 6 Feb 2022 20:10:23 -0500

Quoted 7 lines · Will Cohen

This patch set currently places it in 9p-util only because 9p is the only
place where this issue seems to have come up so far and we were wary of
editing files too far afield, but I have no attachment to its specific
location!

Inline comments are preferred on qemu-devel. Please don't top post ! This complicates the review a lot.

Quoted 7 lines
Quoted 5 lines

This is indeed a good candidate for osdep. This being said, unless there's
some other user in the QEMU code base, it is acceptable to leave it under
9pfs.

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not aware of any work to support any other host OS.

Quoted 7 lines
Quoted 1 line

Cc'ing virtio-fs people for inputs on this topic.

Indeeed, there's a lot of Linux specific code in the virtiofsd - I know people are interested in other platforms, but I'm not sure that's the right starting point.

Dave

Agreeing with Greg here: i.e. I would have placed this into osdep, but I would not insist on it either.

Best regards,
Christian Schoenebeck

author
#25 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 41 lines · Christian Schoenebeck
Quoted 9 lines · On Montag, 7. Februar 2022 11:57:25 CET Dr. David Alan Gilbert wrote:
Quoted 8 lines · * Greg Kurz ([email protected]) wrote:

On Mon, 7 Feb 2022 11:30:18 +0100

Quoted 5 lines · Philippe Mathieu-Daudé
Quoted 4 lines · Greg Kurz

On Sun, 6 Feb 2022 20:10:23 -0500

Quoted 1 line · Will Cohen

This patch set currently places it in 9p-util only because 9p is

the

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

only
place where this issue seems to have come up so far and we were

wary

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 4 lines

of
editing files too far afield, but I have no attachment to its specific
location!

Inline comments are preferred on qemu-devel. Please don't top post

!

Quoted 11 lines
Quoted 11 lines
Quoted 9 lines
Quoted 7 lines

This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's
some other user in the QEMU code base, it is acceptable to leave it under
9pfs.

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not aware

of

Quoted 8 lines
Quoted 3 lines

any work to support any other host OS.

Cc'ing virtio-fs people for inputs on this topic.

Indeeed, there's a lot of Linux specific code in the virtiofsd - I know people are interested in other platforms, but I'm not sure that's the right starting point.

Quoted 10 lines
Quoted 1 line

Dave

Agreeing with Greg here: i.e. I would have placed this into osdep, but I would
not insist on it either.

Best regards,
Christian Schoenebeck

This makes sense. A revised version of this patch, moving qemu_mknodat from 9p-util to osdep and os-posix, is attached below. I'd appreciate any feedback from those looped in here, so that the context isn't lost before resubmitting as a v5 patch, especially since this is starting to touch files outside of 9p.

From c9713c87163da7c96b5357d0d85ac318ae3d3051 Mon Sep 17 00:00:00 2001
From: Keno Fischer <[email protected]>
Date: Sat, 16 Jun 2018 20:56:55 -0400
Subject: [PATCH] 9p: darwin: Implement compatibility for mknodat

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly
- Move qemu_mknodat from 9p-util to osdep and os-posix]

---
 hw/9pfs/9p-local.c   |  4 ++--
 include/qemu/osdep.h | 10 ++++++++++
 os-posix.c           | 34 ++++++++++++++++++++++++++++++++++
 3 files changed, 46 insertions(+), 2 deletions(-)
hw/9pfs/9p-local.c +0 -0
index a0d08e5216..d42ce6d8b8 100644
@@ -682,7 +682,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath

*dir_path,

     if (fs_ctx->export_flags & V9FS_SM_MAPPED ||
         fs_ctx->export_flags & V9FS_SM_MAPPED_FILE) {
-        err = mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
+        err = qemu_mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
         if (err == -1) {
             goto out;
         }
@@ -697,7 +697,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath
*dir_path,
         }
     } else if (fs_ctx->export_flags & V9FS_SM_PASSTHROUGH ||
                fs_ctx->export_flags & V9FS_SM_NONE) {
-        err = mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
+        err = qemu_mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
         if (err == -1) {
             goto out;
         }
include/qemu/osdep.h +0 -0
index d1660d67fa..f3a8367ece 100644
@@ -810,3 +810,13 @@ static inline int

platform_does_not_support_system(const char *command)
#endif

#endif
+
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed. qemu_mknodat is defined in
+ * os-posix.c
+ */
+#ifdef CONFIG_DARWIN
+int pthread_fchdir_np(int fd);
+#endif
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev);

os-posix.c +34 -0
index ae6c9f2a5e..95c1607065 100644
@@ -24,6 +24,7 @@
  */
 
 #include "qemu/osdep.h"
+#include <os/availability.h>
 #include <sys/wait.h>
 #include <pwd.h>
 #include <grp.h>
@@ -332,3 +333,36 @@ int os_mlock(void)
     return -ENOSYS;
 #endif
 }
+
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed.
+ *
+ * Radar filed with Apple for implementing mknodat:
+ * rdar://FB9862426 (https://openradar.appspot.com/FB9862426)
+ */
+#ifdef CONFIG_DARWIN
+
+int pthread_fchdir_np(int fd) API_AVAILABLE(macosx(10.12));
+
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    int preserved_errno, err;
+    if (pthread_fchdir_np(dirfd) < 0) {
+        return -1;
+    }
+    err = mknod(filename, mode, dev);
+    preserved_errno = errno;
+    /* Stop using the thread-local cwd */
+    pthread_fchdir_np(-1);
+    if (err < 0) {
+        errno = preserved_errno;
+    }
+    return err;
+}
+#else
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    return mknodat(dirfd, filename, mode, dev);
+}
+#endif
Signature · 1 line
2.34.1
author
#26 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 185 lines · Will Cohen
Quoted 42 lines · Christian Schoenebeck
Quoted 9 lines · On Montag, 7. Februar 2022 11:57:25 CET Dr. David Alan Gilbert wrote:
Quoted 8 lines · * Greg Kurz ([email protected]) wrote:

On Mon, 7 Feb 2022 11:30:18 +0100

Quoted 5 lines · Philippe Mathieu-Daudé
Quoted 4 lines · Greg Kurz

On Sun, 6 Feb 2022 20:10:23 -0500

Quoted 1 line · Will Cohen

This patch set currently places it in 9p-util only because 9p is

the

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

only
place where this issue seems to have come up so far and we were

wary

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 4 lines

of
editing files too far afield, but I have no attachment to its specific
location!

Inline comments are preferred on qemu-devel. Please don't top

post !

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines

This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's
some other user in the QEMU code base, it is acceptable to leave

it

Quoted 6 lines
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

under
9pfs.

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not aware

of

Quoted 8 lines
Quoted 3 lines

any work to support any other host OS.

Cc'ing virtio-fs people for inputs on this topic.

Indeeed, there's a lot of Linux specific code in the virtiofsd - I know people are interested in other platforms, but I'm not sure that's the right starting point.

Quoted 10 lines
Quoted 1 line

Dave

Agreeing with Greg here: i.e. I would have placed this into osdep, but I would
not insist on it either.

Best regards,
Christian Schoenebeck

This makes sense. A revised version of this patch, moving qemu_mknodat from 9p-util to osdep and os-posix, is attached below. I'd appreciate any feedback from those looped in here, so that the context isn't lost before resubmitting as a v5 patch, especially since this is starting to touch files outside of 9p.

From c9713c87163da7c96b5357d0d85ac318ae3d3051 Mon Sep 17 00:00:00 2001
From: Keno Fischer <[email protected]>
Date: Sat, 16 Jun 2018 20:56:55 -0400
Subject: [PATCH] 9p: darwin: Implement compatibility for mknodat

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly
- Move qemu_mknodat from 9p-util to osdep and os-posix]

---
 hw/9pfs/9p-local.c   |  4 ++--
 include/qemu/osdep.h | 10 ++++++++++
 os-posix.c           | 34 ++++++++++++++++++++++++++++++++++
 3 files changed, 46 insertions(+), 2 deletions(-)
hw/9pfs/9p-local.c +1 -1
index a0d08e5216..d42ce6d8b8 100644
@@ -682,7 +682,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath
 *dir_path,
 
     if (fs_ctx->export_flags & V9FS_SM_MAPPED ||
         fs_ctx->export_flags & V9FS_SM_MAPPED_FILE) {
-        err = mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
+        err = qemu_mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
         if (err == -1) {
             goto out;
         }
hw/9pfs/9p-local.c +1 -1
@@ -697,7 +697,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath
 *dir_path,
         }
     } else if (fs_ctx->export_flags & V9FS_SM_PASSTHROUGH ||
                fs_ctx->export_flags & V9FS_SM_NONE) {
-        err = mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
+        err = qemu_mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
         if (err == -1) {
             goto out;
         }
include/qemu/osdep.h +10 -0
index d1660d67fa..f3a8367ece 100644
@@ -810,3 +810,13 @@ static inline int
 platform_does_not_support_system(const char *command)
 #endif
 
 #endif
+
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed. qemu_mknodat is defined in
+ * os-posix.c
+ */
+#ifdef CONFIG_DARWIN
+int pthread_fchdir_np(int fd);
+#endif
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev);
os-posix.c +34 -1
index ae6c9f2a5e..95c1607065 100644
@@ -24,6 +24,7 @@
  */
 
 #include "qemu/osdep.h"
+#include <os/availability.h>
 #include <sys/wait.h>
 #include <pwd.h>
 #include <grp.h>
@@ -332,3 +333,36 @@ int os_mlock(void)
     return -ENOSYS;
 #endif
 }
+
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed.
+ *
+ * Radar filed with Apple for implementing mknodat:
+ * rdar://FB9862426 (https://openradar.appspot.com/FB9862426)
+ */
+#ifdef CONFIG_DARWIN
+
+int pthread_fchdir_np(int fd) API_AVAILABLE(macosx(10.12));
+
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    int preserved_errno, err;
+    if (pthread_fchdir_np(dirfd) < 0) {
+        return -1;
+    }
+    err = mknod(filename, mode, dev);
+    preserved_errno = errno;
+    /* Stop using the thread-local cwd */
+    pthread_fchdir_np(-1);
+    if (err < 0) {
+        errno = preserved_errno;
+    }
+    return err;
+}
+#else
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    return mknodat(dirfd, filename, mode, dev);
+}
+#endif
--
Quoted 2 lines

2.34.1

Noting on
https://www.qemu.org/docs/master/devel/submitting-a-patch.html#when-resending-patches-add-a-version-tag
that resending a single patch outside of a new entire patch series is not helpful, so I'm resubmitting with this patch substituted in as v5 shortly.

#27 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

In reply to Will Cohen
Quoted 131 lines · On Montag, 7. Februar 2022 22:07:34 CET Will Cohen wrote:
Quoted 50 lines · Christian Schoenebeck
Quoted 9 lines · On Montag, 7. Februar 2022 11:57:25 CET Dr. David Alan Gilbert wrote:
Quoted 8 lines · * Greg Kurz ([email protected]) wrote:

On Mon, 7 Feb 2022 11:30:18 +0100

Quoted 5 lines · Philippe Mathieu-Daudé
Quoted 4 lines · Greg Kurz

On Sun, 6 Feb 2022 20:10:23 -0500

Quoted 1 line · Will Cohen

This patch set currently places it in 9p-util only because 9p is

the

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

only
place where this issue seems to have come up so far and we were

wary

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 4 lines

of
editing files too far afield, but I have no attachment to its specific
location!

Inline comments are preferred on qemu-devel. Please don't top post

!

Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 8 lines

This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's
some other user in the QEMU code base, it is acceptable to leave it
under
9pfs.

Quoted 3 lines
Quoted 3 lines
Quoted 1 line

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not aware

of

Quoted 8 lines
Quoted 3 lines

any work to support any other host OS.

Cc'ing virtio-fs people for inputs on this topic.

Indeeed, there's a lot of Linux specific code in the virtiofsd - I know people are interested in other platforms, but I'm not sure that's the right starting point.

Quoted 8 lines
Quoted 1 line

Dave

Agreeing with Greg here: i.e. I would have placed this into osdep, but I would
not insist on it either.

Best regards,
Christian Schoenebeck

This makes sense. A revised version of this patch, moving qemu_mknodat from 9p-util to osdep and os-posix, is attached below. I'd appreciate any feedback from those looped in here, so that the context isn't lost before resubmitting as a v5 patch, especially since this is starting to touch files outside of 9p.

From c9713c87163da7c96b5357d0d85ac318ae3d3051 Mon Sep 17 00:00:00 2001
From: Keno Fischer <[email protected]>
Date: Sat, 16 Jun 2018 20:56:55 -0400
Subject: [PATCH] 9p: darwin: Implement compatibility for mknodat

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly
- Move qemu_mknodat from 9p-util to osdep and os-posix]

---
 hw/9pfs/9p-local.c   |  4 ++--
 include/qemu/osdep.h | 10 ++++++++++
 os-posix.c           | 34 ++++++++++++++++++++++++++++++++++
 3 files changed, 46 insertions(+), 2 deletions(-)
hw/9pfs/9p-local.c +1 -1
index a0d08e5216..d42ce6d8b8 100644
@@ -682,7 +682,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath
 *dir_path,
 
     if (fs_ctx->export_flags & V9FS_SM_MAPPED ||
         fs_ctx->export_flags & V9FS_SM_MAPPED_FILE) {
-        err = mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
+        err = qemu_mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
         if (err == -1) {
             goto out;
         }
hw/9pfs/9p-local.c +1 -1
@@ -697,7 +697,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath
 *dir_path,
         }
     } else if (fs_ctx->export_flags & V9FS_SM_PASSTHROUGH ||
                fs_ctx->export_flags & V9FS_SM_NONE) {
-        err = mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
+        err = qemu_mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
         if (err == -1) {
             goto out;
         }
include/qemu/osdep.h +0 -0
index d1660d67fa..f3a8367ece 100644
@@ -810,3 +810,13 @@ static inline int
 platform_does_not_support_system(const char *command)
 #endif
 

include/qemu/osdep.h static inline int

Quoted 10 lines
include/qemu/osdep.h +9 -0
 #endif
+
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed. qemu_mknodat is defined in
+ * os-posix.c
+ */
+#ifdef CONFIG_DARWIN
+int pthread_fchdir_np(int fd);
+#endif

I would make that:

#ifdef CONFIG_DARWIN
int pthread_fchdir_np(int fd) API_AVAILABLE(macosx(10.12)); #endif

here and ...

Quoted 18 lines
include/qemu/osdep.h +1 -0
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev);
os-posix.c +2 -0
index ae6c9f2a5e..95c1607065 100644
@@ -24,6 +24,7 @@
  */
 
 #include "qemu/osdep.h"
+#include <os/availability.h>
 #include <sys/wait.h>
 #include <pwd.h>
 #include <grp.h>
@@ -332,3 +333,36 @@ int os_mlock(void)
     return -ENOSYS;
 #endif
 }
+

os-posix.c int os_mlock(void) · line 346

Quoted 10 lines
os-posix.c +10 -0
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed.
+ *
+ * Radar filed with Apple for implementing mknodat:
+ * rdar://FB9862426 (https://openradar.appspot.com/FB9862426)
+ */
+#ifdef CONFIG_DARWIN
+
+int pthread_fchdir_np(int fd) API_AVAILABLE(macosx(10.12));

... drop the duplicate declaration of pthread_fchdir_np() here.

Quoted 22 lines
os-posix.c +22 -0
+
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    int preserved_errno, err;
+    if (pthread_fchdir_np(dirfd) < 0) {
+        return -1;
+    }
+    err = mknod(filename, mode, dev);
+    preserved_errno = errno;
+    /* Stop using the thread-local cwd */
+    pthread_fchdir_np(-1);
+    if (err < 0) {
+        errno = preserved_errno;
+    }
+    return err;
+}
+#else
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t dev)
+{
+    return mknodat(dirfd, filename, mode, dev);
+}
+#endif
author
#28 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

Quoted 212 lines · Christian Schoenebeck
Quoted 5 lines · On Montag, 7. Februar 2022 22:07:34 CET Will Cohen wrote:
Quoted 2 lines · Christian Schoenebeck
Quoted 1 line · On Montag, 7. Februar 2022 11:57:25 CET Dr. David Alan Gilbert wrote:

* Greg Kurz ([email protected]) wrote:

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

On Mon, 7 Feb 2022 11:30:18 +0100

Quoted 5 lines · Philippe Mathieu-Daudé
Quoted 4 lines · Greg Kurz

On Sun, 6 Feb 2022 20:10:23 -0500

Quoted 1 line · Will Cohen

This patch set currently places it in 9p-util only because 9p

is

Quoted 5 lines
Quoted 5 lines

the

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

only
place where this issue seems to have come up so far and we

were

Quoted 9 lines
Quoted 9 lines

wary

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 4 lines

of
editing files too far afield, but I have no attachment to its specific
location!

Inline comments are preferred on qemu-devel. Please don't top

post

Quoted 6 lines
Quoted 6 lines

!

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

This complicates the review a lot.

This is indeed a good candidate for osdep. This being said,

unless

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

there's
some other user in the QEMU code base, it is acceptable to

leave

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

it
under
9pfs.

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not

aware

Quoted 8 lines
Quoted 8 lines

of

Quoted 5 lines
Quoted 3 lines

any work to support any other host OS.

Cc'ing virtio-fs people for inputs on this topic.

Indeeed, there's a lot of Linux specific code in the virtiofsd - I

know

Quoted 6 lines
Quoted 6 lines
Quoted 4 lines

people are interested in other platforms, but I'm not sure that's the right starting point.

Dave

Agreeing with Greg here: i.e. I would have placed this into osdep, but

I

Quoted 7 lines
Quoted 5 lines

would
not insist on it either.

Best regards,
Christian Schoenebeck

This makes sense. A revised version of this patch, moving qemu_mknodat

from

Quoted 68 lines

9p-util to osdep and os-posix, is attached below. I'd appreciate any feedback from those looped in here, so that the context isn't lost before resubmitting as a v5 patch, especially since this is starting to touch files outside of 9p.

From c9713c87163da7c96b5357d0d85ac318ae3d3051 Mon Sep 17 00:00:00 2001
From: Keno Fischer <[email protected]>
Date: Sat, 16 Jun 2018 20:56:55 -0400
Subject: [PATCH] 9p: darwin: Implement compatibility for mknodat

Darwin does not support mknodat. However, to avoid race conditions with later setting the permissions, we must avoid using mknod on the full path instead. We could try to fchdir, but that would cause problems if multiple threads try to call mknodat at the same time. However, luckily there is a solution: Darwin includes a function that sets the cwd for the current thread only.
This should suffice to use mknod safely.

This function (pthread_fchdir_np) is protected by a check in meson in a patch later in tihs series.

[Will Cohen: - Adjust coding style
- Replace clang references with gcc
- Note radar filed with Apple for missing syscall
- Replace direct syscall with pthread_fchdir_np and adjust patch notes accordingly
- Move qemu_mknodat from 9p-util to osdep and os-posix]

---
 hw/9pfs/9p-local.c   |  4 ++--
 include/qemu/osdep.h | 10 ++++++++++
 os-posix.c           | 34 ++++++++++++++++++++++++++++++++++
 3 files changed, 46 insertions(+), 2 deletions(-)
hw/9pfs/9p-local.c +1 -1
index a0d08e5216..d42ce6d8b8 100644
@@ -682,7 +682,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath
 *dir_path,
 
     if (fs_ctx->export_flags & V9FS_SM_MAPPED ||
         fs_ctx->export_flags & V9FS_SM_MAPPED_FILE) {
-        err = mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
+        err = qemu_mknodat(dirfd, name, fs_ctx->fmode | S_IFREG, 0);
         if (err == -1) {
             goto out;
         }
hw/9pfs/9p-local.c +1 -1
@@ -697,7 +697,7 @@ static int local_mknod(FsContext *fs_ctx, V9fsPath
 *dir_path,
         }
     } else if (fs_ctx->export_flags & V9FS_SM_PASSTHROUGH ||
                fs_ctx->export_flags & V9FS_SM_NONE) {
-        err = mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
+        err = qemu_mknodat(dirfd, name, credp->fc_mode, credp->fc_rdev);
         if (err == -1) {
             goto out;
         }
include/qemu/osdep.h +0 -0
index d1660d67fa..f3a8367ece 100644
@@ -810,3 +810,13 @@ static inline int
 platform_does_not_support_system(const char *command)
 #endif
 

include/qemu/osdep.h static inline int

Quoted 10 lines
include/qemu/osdep.h +9 -0
 #endif
+
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed. qemu_mknodat is defined in
+ * os-posix.c
+ */
+#ifdef CONFIG_DARWIN
+int pthread_fchdir_np(int fd);
+#endif

I would make that:

#ifdef CONFIG_DARWIN
int pthread_fchdir_np(int fd) API_AVAILABLE(macosx(10.12)); #endif

here and ...

Quoted 1 line

+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t

dev);

Quoted 17 lines
os-posix.c +2 -0
index ae6c9f2a5e..95c1607065 100644
@@ -24,6 +24,7 @@
  */
 
 #include "qemu/osdep.h"
+#include <os/availability.h>
 #include <sys/wait.h>
 #include <pwd.h>
 #include <grp.h>
@@ -332,3 +333,36 @@ int os_mlock(void)
     return -ENOSYS;
 #endif
 }
+

os-posix.c int os_mlock(void) · line 346

Quoted 10 lines
os-posix.c +10 -0
+/*
+ * As long as mknodat is not available on macOS, this workaround
+ * using pthread_fchdir_np is needed.
+ *
+ * Radar filed with Apple for implementing mknodat:
+ * rdar://FB9862426 (https://openradar.appspot.com/FB9862426)
+ */
+#ifdef CONFIG_DARWIN
+
+int pthread_fchdir_np(int fd) API_AVAILABLE(macosx(10.12));

... drop the duplicate declaration of pthread_fchdir_np() here.

os-posix.c int os_mlock(void) · line 348

Quoted 2 lines
os-posix.c +2 -0
+
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t

dev)

Quoted 6 lines
os-posix.c +6 -0
+{
+    int preserved_errno, err;
+    if (pthread_fchdir_np(dirfd) < 0) {
+        return -1;
+    }
+    err = mknod(filename, mode, dev);

os-posix.c int os_mlock(void)

Quoted 10 lines
os-posix.c +10 -0
+    preserved_errno = errno;
+    /* Stop using the thread-local cwd */
+    pthread_fchdir_np(-1);
+    if (err < 0) {
+        errno = preserved_errno;
+    }
+    return err;
+}
+#else
+int qemu_mknodat(int dirfd, const char *filename, mode_t mode, dev_t

dev)

Quoted 5 lines

os-posix.c int os_mlock(void) · line 368

Quoted 4 lines
os-posix.c +4 -0
+{
+    return mknodat(dirfd, filename, mode, dev);
+}
+#endif

Noted and all of that makes sense to me. Sorry about trying to lump this in v4. If more things come up with v5 I'll stick these changes in for v6 as well. If there's no other major comments with v5, I can either still resubmit as v6 or defer to whatever process is easiest for the maintainers to integrate, given that I accept all of these modifications.

#29 lore raw

Re: [PATCH v4 09/11] 9p: darwin: Implement compatibility for mknodat

In reply to Greg Kurz
Quoted 20 lines · Greg Kurz
Quoted 17 lines · On Mon, 7 Feb 2022 11:30:18 +0100 Philippe Mathieu-Daudé <[email protected]>…
Quoted 15 lines · Greg Kurz
Quoted 5 lines · On Sun, 6 Feb 2022 20:10:23 -0500 Will Cohen <[email protected]> wrote:

This patch set currently places it in 9p-util only because 9p is the only place where this issue seems to have come up so far and we were wary of editing files too far afield, but I have no attachment to its specific location!

Inline comments are preferred on qemu-devel. Please don't top post ! This complicates the review a lot.

This is indeed a good candidate for osdep. This being said, unless there's some other user in the QEMU code base, it is acceptable to leave it under 9pfs.

Quoted 5 lines
Quoted 1 line

virtiofsd could eventually use it.

Indeed but virtiofsd is for linux hosts only AFAICT and I'm not aware of any work to support any other host OS.

[ CC Sergio ]

Will like to support virtiofs on other host OS. Getting rid of Linux specific parts should be doable. I think bigger challenge is how to make vhost-user stuff work on other OS, like macOS.

If virtiofsd was somehow running as part of qemu (and not as a separate process), then making rest of the filesystem code to work on other OS should not be too hard, I guess.

So question is, can one somehow run same virtiofsd code both as part of qemu as well as separate daemon based on need (and one does not have to maintain two separate code bases).

Thanks
Vivek

Quoted 3 lines

Cc'ing virtio-fs people for inputs on this topic.

author
#30 lore raw

PATCH v4 10/11 9p: darwin: meson: Allow VirtFS on Darwin

In reply to Will Cohen

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]
---
 fsdev/meson.build |  1 +
 meson.build       | 14 ++++++++++----
 2 files changed, 11 insertions(+), 4 deletions(-)
fsdev/meson.build +1 -0
index adf57cc43e..b632b66348 100644
@@ -7,6 +7,7 @@ fsdev_ss.add(when: ['CONFIG_FSDEV_9P'], if_true: files(
   'qemu-fsdev.c',
 ), if_false: files('qemu-fsdev-dummy.c'))
 softmmu_ss.add_all(when: 'CONFIG_LINUX', if_true: fsdev_ss)
+softmmu_ss.add_all(when: 'CONFIG_DARWIN', if_true: fsdev_ss)
 
 if have_virtfs_proxy_helper
   executable('virtfs-proxy-helper',
meson.build +10 -4
index 5f43355071..6b4adf7e15 100644
@@ -1421,17 +1421,23 @@ if not get_option('dbus_display').disabled()
   endif
 endif
 
-have_virtfs = (targetos == 'linux' and
+if targetos == 'darwin' and cc.has_function('pthread_fchdir_np')
+  have_virtfs = have_system
+else
+  have_virtfs = (targetos == 'linux' and
     have_system and
     libattr.found() and
     libcap_ng.found())
+endif
 
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and have_tools
 
 if get_option('virtfs').enabled()
   if not have_virtfs
-    if targetos != 'linux'
-      error('virtio-9p (virtfs) requires Linux')
+    if targetos != 'linux' and targetos != 'darwin'
+      error('virtio-9p (virtfs) requires Linux or Darwin')
+    elif targetos == 'darwin' and not cc.has_function('pthread_fchdir_np')
+      error('virtio-9p (virtfs) on Darwin requires the presence of pthread_fchdir_np')
     elif not libcap_ng.found() or not libattr.found()
       error('virtio-9p (virtfs) requires libcap-ng-devel and libattr-devel')
     elif not have_system
Signature · 1 line
2.32.0 (Apple Git-132)
#31 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

Quoted 15 lines · Will Cohen

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]
---
  fsdev/meson.build |  1 +
  meson.build       | 14 ++++++++++----
  2 files changed, 11 insertions(+), 4 deletions(-)

meson.build if not get_option('dbus_display').disabled() · line 1433

Quoted 2 lines
meson.build +1 -1
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and have_tools

Why do you restrict the proxy-helper to Linux?

author
#32 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

Only because porting the proxy-helper to macOS is outside the scope of this particular patch. While some initial concepts around it have been considered by some of the contributors to this patch, those implementations weren't tested enough and the security implications weren't considered in full. We assume that this could be an additional implementation later on, if the functionality is considered important down the road.

Quoted 22 lines · Philippe Mathieu-Daudé
Quoted 15 lines · Will Cohen

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]
---
  fsdev/meson.build |  1 +
  meson.build       | 14 ++++++++++----
  2 files changed, 11 insertions(+), 4 deletions(-)

Review

Quoted 2 lines
(quoted diff) +1 -1
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

Why do you restrict the proxy-helper to Linux?

#33 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

Quoted 25 lines · On Montag, 7. Februar 2022 02:05:32 CET Will Cohen wrote:

On Sun, Feb 6, 2022 at 4:22 PM Philippe Mathieu-Daudé <[email protected]>

Quoted 22 lines · wrote:
Quoted 16 lines · Will Cohen

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]

---

  fsdev/meson.build |  1 +
  meson.build       | 14 ++++++++++----
  2 files changed, 11 insertions(+), 4 deletions(-)

Review

Quoted 2 lines
(quoted diff) +1 -1
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

Quoted 8 lines
Quoted 1 line

Why do you restrict the proxy-helper to Linux?

Only because porting the proxy-helper to macOS is outside the scope of this particular patch. While some initial concepts around it have been considered by some of the contributors to this patch, those implementations weren't tested enough and the security implications weren't considered in full. We assume that this could be an additional implementation later on, if the functionality is considered important down the road.

In general that's fine with me. I would have probably made that "targetos != 'darwin'" instead of "targetos == 'linux'", but I leave that up to you.

On the long term we will probably deprecate the 9p 'proxy' fs driver anyway. While it had some good ideas, being realistic though: nobody has worked on the 9p proxy driver/backend for many years and it is not in good shape.

I can imagine that due to the ground being laid by these series, that we will also open 9p for BSD, but that should be done a bit later and hence does not belong into these series.

But once again: it would not have hurt to make your intentions clear either in the commit log or by in-source comment. :)

Best regards,
Christian Schoenebeck

author
#34 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

Quoted 54 lines · Christian Schoenebeck
Quoted 20 lines · On Montag, 7. Februar 2022 02:05:32 CET Will Cohen wrote:

On Sun, Feb 6, 2022 at 4:22 PM Philippe Mathieu-Daudé <[email protected]>

Quoted 17 lines · wrote:
Quoted 16 lines · Will Cohen

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]

---

  fsdev/meson.build |  1 +
  meson.build       | 14 ++++++++++----
  2 files changed, 11 insertions(+), 4 deletions(-)
Quoted 8 lines
Quoted 6 lines

Review

Quoted 2 lines
(quoted diff) +1 -1
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

Why do you restrict the proxy-helper to Linux?

Only because porting the proxy-helper to macOS is outside the scope of

this

Quoted 2 lines

particular patch. While some initial concepts around it have been considered by some of the contributors to this patch, those

implementations

Quoted 3 lines

weren't tested enough and the security implications weren't considered in full. We assume that this could be an additional implementation later on, if the functionality is considered important down the road.

In general that's fine with me. I would have probably made that "targetos != 'darwin'" instead of "targetos == 'linux'", but I leave that up
to you.

On the long term we will probably deprecate the 9p 'proxy' fs driver anyway.
While it had some good ideas, being realistic though: nobody has worked on the
9p proxy driver/backend for many years and it is not in good shape.

I can imagine that due to the ground being laid by these series, that we will
also open 9p for BSD, but that should be done a bit later and hence does not
belong into these series.

Quoted 9 lines

But once again: it would not have hurt to make your intentions clear either in
the commit log or by in-source comment. :)

Best regards,
Christian Schoenebeck

Acknowledged! For v5 will change it to != 'darwin' and note it as well, for clarity.

#35 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

In reply to Christian Schoenebeck
Quoted 35 lines · On Mon, 07 Feb 2022 15:15:46 +0100 Christian Schoenebeck…
Quoted 25 lines · On Montag, 7. Februar 2022 02:05:32 CET Will Cohen wrote:

On Sun, Feb 6, 2022 at 4:22 PM Philippe Mathieu-Daudé <[email protected]>

Quoted 22 lines · wrote:
Quoted 16 lines · Will Cohen

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]

---

  fsdev/meson.build |  1 +
  meson.build       | 14 ++++++++++----
  2 files changed, 11 insertions(+), 4 deletions(-)

Review

Quoted 2 lines
(quoted diff) +1 -1
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

Quoted 8 lines
Quoted 1 line

Why do you restrict the proxy-helper to Linux?

Only because porting the proxy-helper to macOS is outside the scope of this particular patch. While some initial concepts around it have been considered by some of the contributors to this patch, those implementations weren't tested enough and the security implications weren't considered in full. We assume that this could be an additional implementation later on, if the functionality is considered important down the road.

Quoted 8 lines

In general that's fine with me. I would have probably made that "targetos != 'darwin'" instead of "targetos == 'linux'", but I leave that up to you.

On the long term we will probably deprecate the 9p 'proxy' fs driver anyway. While it had some good ideas, being realistic though: nobody has worked on the 9p proxy driver/backend for many years and it is not in good shape.

It definitely isn't indeed. Also it is super slow by design since the round trip of a 9p request involves QEMU on both entry and exit:

[guest] --> [QEMU]--> [virtfs-proxy-helper]-->[QEMU]-->[guest]

A more modern and efficient approach would be to have a vhost-user-9p implementation : requests would be directly handled by the external process, without QEMU hops. But this a fair amount of work.

Quoted 11 lines

I can imagine that due to the ground being laid by these series, that we will also open 9p for BSD, but that should be done a bit later and hence does not belong into these series.

But once again: it would not have hurt to make your intentions clear either in the commit log or by in-source comment. :)

Best regards,
Christian Schoenebeck

#36 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

Quoted 54 lines · On Montag, 7. Februar 2022 15:39:30 CET Greg Kurz wrote:

On Mon, 07 Feb 2022 15:15:46 +0100

Quoted 39 lines · Christian Schoenebeck
Quoted 27 lines · On Montag, 7. Februar 2022 02:05:32 CET Will Cohen wrote:

On Sun, Feb 6, 2022 at 4:22 PM Philippe Mathieu-Daudé <[email protected]>

Quoted 17 lines · wrote:
Quoted 16 lines · Will Cohen

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]

---

  fsdev/meson.build |  1 +
  meson.build       | 14 ++++++++++----
  2 files changed, 11 insertions(+), 4 deletions(-)
Quoted 6 lines

Review

Quoted 2 lines
(quoted diff) +1 -1
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

Why do you restrict the proxy-helper to Linux?

Quoted 10 lines

Only because porting the proxy-helper to macOS is outside the scope of this
particular patch. While some initial concepts around it have been considered by some of the contributors to this patch, those implementations
weren't tested enough and the security implications weren't considered in
full. We assume that this could be an additional implementation later on,
if the functionality is considered important down the road.

Quoted 8 lines

In general that's fine with me. I would have probably made that "targetos != 'darwin'" instead of "targetos == 'linux'", but I leave that up to you.

On the long term we will probably deprecate the 9p 'proxy' fs driver anyway. While it had some good ideas, being realistic though: nobody has worked on the 9p proxy driver/backend for many years and it is not in good shape.

It definitely isn't indeed. Also it is super slow by design since the round trip of a 9p request involves QEMU on both entry and exit:

Quoted 5 lines
   [guest] --> [QEMU]--> [virtfs-proxy-helper]-->[QEMU]-->[guest]

A more modern and efficient approach would be to have a vhost-user-9p implementation : requests would be directly handled by the external process, without QEMU hops. But this a fair amount of work.

That's already a bit offtopic, but how would you imagine that to work? You mean a system dependent solution that e.g. plugs in into KVM or something?

Quoted 9 lines
Quoted 9 lines

I can imagine that due to the ground being laid by these series, that we will also open 9p for BSD, but that should be done a bit later and hence does not belong into these series.

But once again: it would not have hurt to make your intentions clear either in the commit log or by in-source comment. :)

Best regards,
Christian Schoenebeck

#37 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

In reply to Will Cohen
Quoted 32 lines · On Sonntag, 6. Februar 2022 21:07:18 CET Will Cohen wrote:

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]
---
 fsdev/meson.build |  1 +
 meson.build       | 14 ++++++++++----
 2 files changed, 11 insertions(+), 4 deletions(-)
fsdev/meson.build +1 -0
index adf57cc43e..b632b66348 100644
@@ -7,6 +7,7 @@ fsdev_ss.add(when: ['CONFIG_FSDEV_9P'], if_true: files(
   'qemu-fsdev.c',
 ), if_false: files('qemu-fsdev-dummy.c'))
 softmmu_ss.add_all(when: 'CONFIG_LINUX', if_true: fsdev_ss)
+softmmu_ss.add_all(when: 'CONFIG_DARWIN', if_true: fsdev_ss)
 
 if have_virtfs_proxy_helper
   executable('virtfs-proxy-helper',
meson.build +0 -0
index 5f43355071..6b4adf7e15 100644

meson.build if not get_option('dbus_display').disabled() · line 1425

Quoted 6 lines
meson.build +2 -1
@@ -1421,17 +1421,23 @@ if not get_option('dbus_display').disabled()
   endif
 endif
 
-have_virtfs = (targetos == 'linux' and
+if targetos == 'darwin' and cc.has_function('pthread_fchdir_np')
+  have_virtfs = have_system

As you are going for a v5 anyway: I would add an error message here if pthread_fchdir_np() is not available. Because it is a bit frustrating for users if their options silently got ignored without any indication why.

Quoted 22 lines
meson.build +4 -1
+else
+  have_virtfs = (targetos == 'linux' and
     have_system and
     libattr.found() and
     libcap_ng.found())
+endif
 
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

if get_option('virtfs').enabled()
if not have_virtfs

meson.build +4 -2
-    if targetos != 'linux'
-      error('virtio-9p (virtfs) requires Linux')
+    if targetos != 'linux' and targetos != 'darwin'
+      error('virtio-9p (virtfs) requires Linux or Darwin')
+    elif targetos == 'darwin' and not cc.has_function('pthread_fchdir_np')
+      error('virtio-9p (virtfs) on Darwin requires the presence of

pthread_fchdir_np') elif not libcap_ng.found() or not libattr.found()
error('virtio-9p (virtfs) requires libcap-ng-devel and
libattr-devel') elif not have_system

author
#38 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

Quoted 55 lines · Christian Schoenebeck
Quoted 32 lines · On Sonntag, 6. Februar 2022 21:07:18 CET Will Cohen wrote:

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]
---
 fsdev/meson.build |  1 +
 meson.build       | 14 ++++++++++----
 2 files changed, 11 insertions(+), 4 deletions(-)
fsdev/meson.build +1 -0
index adf57cc43e..b632b66348 100644
@@ -7,6 +7,7 @@ fsdev_ss.add(when: ['CONFIG_FSDEV_9P'], if_true: files(
   'qemu-fsdev.c',
 ), if_false: files('qemu-fsdev-dummy.c'))
 softmmu_ss.add_all(when: 'CONFIG_LINUX', if_true: fsdev_ss)
+softmmu_ss.add_all(when: 'CONFIG_DARWIN', if_true: fsdev_ss)
 
 if have_virtfs_proxy_helper
   executable('virtfs-proxy-helper',
meson.build +0 -0
index 5f43355071..6b4adf7e15 100644

meson.build if not get_option('dbus_display').disabled() · line 1425

Quoted 6 lines
meson.build +2 -1
@@ -1421,17 +1421,23 @@ if not get_option('dbus_display').disabled()
   endif
 endif
 
-have_virtfs = (targetos == 'linux' and
+if targetos == 'darwin' and cc.has_function('pthread_fchdir_np')
+  have_virtfs = have_system

As you are going for a v5 anyway: I would add an error message here if pthread_fchdir_np() is not available. Because it is a bit frustrating for users if their options silently got ignored without any indication why.

Quoted 11 lines
meson.build +4 -1
+else
+  have_virtfs = (targetos == 'linux' and
     have_system and
     libattr.found() and
     libcap_ng.found())
+endif
 
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

Quoted 10 lines

meson.build if not get_option('dbus_display').disabled()

Quoted 7 lines

if get_option('virtfs').enabled()
if not have_virtfs

meson.build +3 -2
-    if targetos != 'linux'
-      error('virtio-9p (virtfs) requires Linux')
+    if targetos != 'linux' and targetos != 'darwin'
+      error('virtio-9p (virtfs) requires Linux or Darwin')
+    elif targetos == 'darwin' and not

cc.has_function('pthread_fchdir_np')

Quoted 2 lines

+ error('virtio-9p (virtfs) on Darwin requires the presence of
pthread_fchdir_np')

Does the error message here suffice for that need? Right now if they're running a system without pthread_fchdir_np and don't specify the option, I think it'll just quietly disable, but if they --enable-virtfs and the function isn't there, they should get a note. I assume this is better, so that the ability to compile isn't contingent on having the latest OS, even if full support for older OSes isn't provided.

Quoted 6 lines

elif not libcap_ng.found() or not libattr.found()

Quoted 2 lines

error('virtio-9p (virtfs) requires libcap-ng-devel and
libattr-devel') elif not have_system

#39 lore raw

Re: [PATCH v4 10/11] 9p: darwin: meson: Allow VirtFS on Darwin

Quoted 83 lines · On Montag, 7. Februar 2022 15:37:00 CET Will Cohen wrote:
Quoted 69 lines · Christian Schoenebeck
Quoted 38 lines · On Sonntag, 6. Februar 2022 21:07:18 CET Will Cohen wrote:

From: Keno Fischer <[email protected]>

[Michael Roitzsch: - Rebase for NixOS]
[Will Cohen: - Rebase to master]
[Will Cohen: - Add check for pthread_fchdir_np to virtfs]

---

 fsdev/meson.build |  1 +
 meson.build       | 14 ++++++++++----
 2 files changed, 11 insertions(+), 4 deletions(-)
fsdev/meson.build +1 -0
index adf57cc43e..b632b66348 100644
@@ -7,6 +7,7 @@ fsdev_ss.add(when: ['CONFIG_FSDEV_9P'], if_true: files(
 
   'qemu-fsdev.c',
 
 ), if_false: files('qemu-fsdev-dummy.c'))
 softmmu_ss.add_all(when: 'CONFIG_LINUX', if_true: fsdev_ss)
 
+softmmu_ss.add_all(when: 'CONFIG_DARWIN', if_true: fsdev_ss)

if have_virtfs_proxy_helper

executable('virtfs-proxy-helper',

meson.build +0 -0
index 5f43355071..6b4adf7e15 100644

meson.build if not get_option('dbus_display').disabled() · line 1427

Quoted 8 lines
meson.build +2 -1
@@ -1421,17 +1421,23 @@ if not get_option('dbus_display').disabled()
 
   endif
 
 endif
 
-have_virtfs = (targetos == 'linux' and
+if targetos == 'darwin' and cc.has_function('pthread_fchdir_np')
+  have_virtfs = have_system

As you are going for a v5 anyway: I would add an error message here if pthread_fchdir_np() is not available. Because it is a bit frustrating for users if their options silently got ignored without any indication why.

Quoted 17 lines
meson.build +4 -1
+else
+  have_virtfs = (targetos == 'linux' and
 
     have_system and
     libattr.found() and
     libcap_ng.found())
 
+endif
 
-have_virtfs_proxy_helper = have_virtfs and have_tools
+have_virtfs_proxy_helper = targetos == 'linux' and have_virtfs and

have_tools

if get_option('virtfs').enabled()

if not have_virtfs

Quoted 10 lines

meson.build if not get_option('dbus_display').disabled()

Quoted 5 lines
meson.build +3 -2
-    if targetos != 'linux'
-      error('virtio-9p (virtfs) requires Linux')
+    if targetos != 'linux' and targetos != 'darwin'
+      error('virtio-9p (virtfs) requires Linux or Darwin')
+    elif targetos == 'darwin' and not

cc.has_function('pthread_fchdir_np')

Quoted 2 lines

+ error('virtio-9p (virtfs) on Darwin requires the presence of
pthread_fchdir_np')

Quoted 6 lines

Does the error message here suffice for that need? Right now if they're running a system without pthread_fchdir_np and don't specify the option, I think it'll just quietly disable, but if they --enable-virtfs and the function isn't there, they should get a note. I assume this is better, so that the ability to compile isn't contingent on having the latest OS, even if full support for older OSes isn't provided.

Ah, got it. Yes, makes sense.

But what I would definitely change is the precise error message text here:
"Darwin" is a bit awkward for a regular user, because most macOS users never heard of "Darwin" in the context of Apple systems before. Using the term "darwin" in code is fine as it can be assumed that developers know the background, but as for regular users I would make it more clear that this is actually about macOS, e.g:

error('virtio-9p (virtfs) requires either Linux or Darwin (macOS)')

I don't mind how to write that exactly; braces, slash, replacing Darwin by macOS or whatever, but it should mention 'macOS' here in some form.

Quoted 5 lines
Quoted 5 lines

elif not libcap_ng.found() or not libattr.found()

Quoted 3 lines

error('virtio-9p (virtfs) requires libcap-ng-devel and

libattr-devel') elif not have_system

author
#40 lore raw

PATCH v4 11/11 9p: darwin: Adjust assumption on virtio-9p-test

In reply to Will Cohen

The previous test depended on the assumption that P9_DOTL_AT_REMOVEDIR and AT_REMOVEDIR have the same value.

While this is true on Linux, it is not true everywhere, and leads to an incorrect test failure on unlink_at, noticed when adding 9p to darwin:

Received response 7 (RLERROR) instead of 77 (RUNLINKAT) Rlerror has errno 22 (Invalid argument)
**

ERROR:../tests/qtest/virtio-9p-test.c:305:v9fs_req_recv: assertion failed (hdr.id == id): (7 == 77) Bail out!

ERROR:../tests/qtest/virtio-9p-test.c:305:v9fs_req_recv: assertion failed (hdr.id == id): (7 == 77)

[Will Cohen: - Add explanation of patch and description
of pre-patch test failure]

---
 tests/qtest/virtio-9p-test.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tests/qtest/virtio-9p-test.c +1 -1
index 41fed41de1..6bcf89f0f8 100644
@@ -1270,7 +1270,7 @@ static void fs_unlinkat_dir(void *obj, void *data, QGuestAllocator *t_alloc)
     /* ... and is actually a directory */
     g_assert((st.st_mode & S_IFMT) == S_IFDIR);
 
-    do_unlinkat(v9p, "/", "02", AT_REMOVEDIR);
+    do_unlinkat(v9p, "/", "02", P9_DOTL_AT_REMOVEDIR);
     /* directory should be gone now */
     g_assert(stat(new_dir, &st) != 0);
 
Signature · 1 line
2.32.0 (Apple Git-132)
Acked-by on v4 11/11
#41 lore raw

Re: [PATCH v4 11/11] 9p: darwin: Adjust assumption on virtio-9p-test

Quoted 29 lines · Will Cohen

The previous test depended on the assumption that P9_DOTL_AT_REMOVEDIR and AT_REMOVEDIR have the same value.

While this is true on Linux, it is not true everywhere, and leads to an incorrect test failure on unlink_at, noticed when adding 9p to darwin:

Received response 7 (RLERROR) instead of 77 (RUNLINKAT) Rlerror has errno 22 (Invalid argument)
**

ERROR:../tests/qtest/virtio-9p-test.c:305:v9fs_req_recv: assertion failed (hdr.id == id): (7 == 77) Bail out!

ERROR:../tests/qtest/virtio-9p-test.c:305:v9fs_req_recv: assertion failed (hdr.id == id): (7 == 77)

[Will Cohen: - Add explanation of patch and description
of pre-patch test failure]

---
  tests/qtest/virtio-9p-test.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
tests/qtest/virtio-9p-test.c +0 -0
index 41fed41de1..6bcf89f0f8 100644

tests/qtest/virtio-9p-test.c static void fs_unlinkat_dir(void *obj, void *data, QGuestAllocator *t_alloc) · line 1276

Quoted 8 lines
tests/qtest/virtio-9p-test.c +1 -1
@@ -1270,7 +1270,7 @@ static void fs_unlinkat_dir(void *obj, void *data, QGuestAllocator *t_alloc)
      /* ... and is actually a directory */
      g_assert((st.st_mode & S_IFMT) == S_IFDIR);
 
-    do_unlinkat(v9p, "/", "02", AT_REMOVEDIR);
+    do_unlinkat(v9p, "/", "02", P9_DOTL_AT_REMOVEDIR);
      /* directory should be gone now */
      g_assert(stat(new_dir, &st) != 0);
 
#42 lore raw
In reply to Will Cohen
Quoted 6 lines · On Sonntag, 6. Februar 2022 21:07:08 CET Will Cohen wrote:

This is a followup to
https://lists.nongnu.org/archive/html/qemu-devel/2022-01/msg05993.html,
adding 9p server support for Darwin.

Since v3, the following changes have been made:

Quoted 13 lines

- Move XATTR_SIZE_MAX to P9_XATTR_SIZE MAX in 9p.h, and provide explanatory context as preliminary solution - Add explanatory note surrounding virtio-9p-test with output of pre-patch failing test - Remove superfluous header guards from file-opt-9p
- Add note about virtfs-proxy-helper being disabled on non-linux for this patch series - Note radar filed with Apple for missing mknodat syscall
- Replace direct syscall to pthread_fchdir with pthread_fchdir_np, and add check for this function’s presence in meson - Ensure that d_seekoff is filled using telldir on darwin, and create qemu_dirent_off helper to decide which to access. - Ensure that [amc]tim.tv_sec are all initialized alongside [amc]tim.tv_nsec in 9p-proxy - Ensure that all patch email addresses are valid
- Add telldir error handling for dirent on darwin

As this series already has seen some revisions and is on a good way to become queued soon: it is helpful to immediately see here which patches exactly got changed, because some of the patches already look fine.

For already reviewed patches that you won't change: you can take over people's reviewed-by tags in the next revision.

Best regards,
Christian Schoenebeck

Quoted 33 lines

Keno Fischer (10):
9p: linux: Fix a couple Linux assumptions 9p: Rename 9p-util -> 9p-util-linux
9p: darwin: Handle struct stat(fs) differences 9p: darwin: Handle struct dirent differences 9p: darwin: Ignore O_{NOATIME, DIRECT}
9p: darwin: Move XATTR_SIZE_MAX->P9_XATTR_SIZE_MAX 9p: darwin: *xattr_nofollow implementations 9p: darwin: Compatibility for f/l*xattr 9p: darwin: Implement compatibility for mknodat 9p: darwin: meson: Allow VirtFS on Darwin

Will Cohen (1):
9p: darwin: Adjust assumption on virtio-9p-test

 fsdev/file-op-9p.h                     |  9 ++-
 fsdev/meson.build                      |  1 +
 hw/9pfs/9p-local.c                     | 28 ++++++--
 hw/9pfs/9p-proxy.c                     | 38 ++++++++++-
 hw/9pfs/9p-synth.c                     |  6 ++
 hw/9pfs/9p-util-darwin.c               | 91 ++++++++++++++++++++++++++
 hw/9pfs/{9p-util.c => 9p-util-linux.c} |  7 +-
 hw/9pfs/9p-util.h                      | 38 +++++++++++
 hw/9pfs/9p.c                           | 50 ++++++++++++--
 hw/9pfs/9p.h                           | 11 ++++
 hw/9pfs/codir.c                        |  7 ++
 hw/9pfs/meson.build                    |  3 +-
 include/qemu/xattr.h                   |  4 +-
 meson.build                            | 14 ++--
 tests/qtest/virtio-9p-test.c           |  2 +-
 15 files changed, 285 insertions(+), 24 deletions(-)
 create mode 100644 hw/9pfs/9p-util-darwin.c
 rename hw/9pfs/{9p-util.c => 9p-util-linux.c} (90%)
author
#43 lore raw
Quoted 30 lines · Christian Schoenebeck
Quoted 7 lines · On Sonntag, 6. Februar 2022 21:07:08 CET Will Cohen wrote:

This is a followup to
https://lists.nongnu.org/archive/html/qemu-devel/2022-01/msg05993.html,
adding 9p server support for Darwin.

Since v3, the following changes have been made:

- Move XATTR_SIZE_MAX to P9_XATTR_SIZE MAX in 9p.h, and provide

explanatory

Quoted 6 lines

context as preliminary solution - Add explanatory note surrounding virtio-9p-test with output of pre-patch failing test - Remove superfluous header guards from file-opt-9p
- Add note about virtfs-proxy-helper being disabled on non-linux for this patch series - Note radar filed with Apple for missing mknodat syscall
- Replace direct syscall to pthread_fchdir with pthread_fchdir_np, and

add

Quoted 2 lines

check for this function’s presence in meson - Ensure that d_seekoff is filled using telldir on darwin, and create qemu_dirent_off helper to

decide

Quoted 4 lines

which to access. - Ensure that [amc]tim.tv_sec are all initialized alongside [amc]tim.tv_nsec in 9p-proxy - Ensure that all patch email addresses are valid
- Add telldir error handling for dirent on darwin

As this series already has seen some revisions and is on a good way to become
queued soon: it is helpful to immediately see here which patches exactly got
changed, because some of the patches already look fine.

Quoted 42 lines

For already reviewed patches that you won't change: you can take over people's
reviewed-by tags in the next revision.

Best regards,
Christian Schoenebeck

Quoted 33 lines

Keno Fischer (10):
9p: linux: Fix a couple Linux assumptions 9p: Rename 9p-util -> 9p-util-linux
9p: darwin: Handle struct stat(fs) differences 9p: darwin: Handle struct dirent differences 9p: darwin: Ignore O_{NOATIME, DIRECT}
9p: darwin: Move XATTR_SIZE_MAX->P9_XATTR_SIZE_MAX 9p: darwin: *xattr_nofollow implementations 9p: darwin: Compatibility for f/l*xattr 9p: darwin: Implement compatibility for mknodat 9p: darwin: meson: Allow VirtFS on Darwin

Will Cohen (1):
9p: darwin: Adjust assumption on virtio-9p-test

 fsdev/file-op-9p.h                     |  9 ++-
 fsdev/meson.build                      |  1 +
 hw/9pfs/9p-local.c                     | 28 ++++++--
 hw/9pfs/9p-proxy.c                     | 38 ++++++++++-
 hw/9pfs/9p-synth.c                     |  6 ++
 hw/9pfs/9p-util-darwin.c               | 91 ++++++++++++++++++++++++++
 hw/9pfs/{9p-util.c => 9p-util-linux.c} |  7 +-
 hw/9pfs/9p-util.h                      | 38 +++++++++++
 hw/9pfs/9p.c                           | 50 ++++++++++++--
 hw/9pfs/9p.h                           | 11 ++++
 hw/9pfs/codir.c                        |  7 ++
 hw/9pfs/meson.build                    |  3 +-
 include/qemu/xattr.h                   |  4 +-
 meson.build                            | 14 ++--
 tests/qtest/virtio-9p-test.c           |  2 +-
 15 files changed, 285 insertions(+), 24 deletions(-)
 create mode 100644 hw/9pfs/9p-util-darwin.c
 rename hw/9pfs/{9p-util.c => 9p-util-linux.c} (90%)

Understood! Previous reviewed-by tags are already in v4, and these new ones will go into v5!

For reference in terms of changes, patches that were touched in v4:

Keno Fischer (10):
9p: linux: Fix a couple Linux assumptions (1/11) 9p: darwin: Handle struct stat(fs) differences (3/11) 9p: darwin: Handle struct dirent differences (4/11) 9p: darwin: Move XATTR_SIZE_MAX->P9_XATTR_SIZE_MAX (5/11) 9p: darwin: Implement compatibility for mknodat (9/11) 9p: darwin: meson: Allow VirtFS on Darwin (10/11)

Will Cohen (1):
9p: darwin: Adjust assumption on virtio-9p-test (11/11)

Additional changes to patches will be highlighted in the opening note for v5.