PATCH: add --no-symlink-perms option for ignoring symlink permissions on transfer
RolandK via rsync <[email protected]> Sun, 31 May 2026 14:16:15 +0200
| Newsgroups | gmane.network.rsync.general |
|---|---|
| Message-ID | <[email protected]> |
This is a multi-part message in MIME format.
--===============5743251892575364950==
Content-Type: multipart/alternative;
boundary="------------v0ICkeHnT5hDvQerAq0L1fDi"
This is a multi-part message in MIME format.
--------------v0ICkeHnT5hDvQerAq0L1fDi
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable
hello,
i have made a patch , adding an option for for avoiding=20
(bulk-)recreation of symlinks when transferring from macos to linux for=20
example.
on linux (and possibly other platforms) , symlinks have generic 777=20
permission.
on macos, you can set individual permission with chmod -h.
patch is at at=20
https://github.com/devZer0/rsync-workingcopy/commit/7a8d295e531627e0a3d9e4=
1ceedf10ab9be7fb5e
we have checking for lchmod() in autoconf, but for lchmod there seems=20
stub in linux glibc headers which makes this check return success (where=
=20
it shouldn't)
i spent some thoughts if we better do compile time check for platform to=
=20
decide to ignore that or not ,=C2=A0 or do compile time or runtime check f=
or=20
lchmod() syscall return values, but i think it's more easy and=20
predictable, to control this manually.=C2=A0 it's=C2=A0 more transparent a=
nd less=20
magic this way.
also see:
https://github.com/RsyncProject/rsync/issues/443
maybe somebody want's to review/test and provide feedback.
regards
Roland
ps:
and yes, i am also using AI. i'm a sysadmin and not a programmer.
--------------v0ICkeHnT5hDvQerAq0L1fDi
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE html>
<html>
<head>
<meta http-equiv=3D"content-type" content=3D"text/html; charset=3DUTF-=
8">
</head>
<body>
<p>hello,=C2=A0<br>
<br>
i have made a patch , adding an option for for avoiding
(bulk-)recreation of symlinks when transferring from macos to
linux for example.=C2=A0<br>
<br>
on linux (and possibly other platforms) , symlinks have generic
777 permission.=C2=A0<br>
on macos, you can set individual permission with chmod -h.=C2=A0<br>
<br>
patch is at at
<a class=3D"moz-txt-link-freetext" href=3D"https://github.com/devZer0/rsyn=
c-workingcopy/commit/7a8d295e531627e0a3d9e41ceedf10ab9be7fb5e">https://git=
hub.com/devZer0/rsync-workingcopy/commit/7a8d295e531627e0a3d9e41ceedf10ab9=
be7fb5e</a><br>
<br>
we have checking for lchmod() in autoconf, but for lchmod there
seems stub in linux glibc headers which makes this check return
success (where it shouldn't)=C2=A0<br>
<br>
i spent some thoughts if we better do compile time check for
platform to decide to ignore that or not ,=C2=A0 or do compile time =
or
runtime check for=C2=A0<span></span><span class=3D"=CD=BCm">lchmod</=
span><span>()
syscall return values, but i think it's more easy and
predictable, to control this manually.=C2=A0 it's=C2=A0 more=C2=A0=
</span>transparent
and less magic this way.=C2=A0<br>
<br>
also see:<br>
<a class=3D"moz-txt-link-freetext" href=3D"https://github.com/RsyncP=
roject/rsync/issues/443">https://github.com/RsyncProject/rsync/issues/443<=
/a><br>
<br>
maybe somebody want's to review/test and provide feedback.<br>
<br>
regards<br>
Roland<br>
<br>
ps:<br>
and yes, i am also using AI. i'm a sysadmin and not a programmer.=C2=
=A0</p>
</body>
</html>
--------------v0ICkeHnT5hDvQerAq0L1fDi--
--===============5743251892575364950==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
--
Please use reply-all for most replies to avoid omitting the mailing list.
To unsubscribe or change options: https://lists.samba.org/mailman/listinfo/rsync
Before posting, read: http://www.catb.org/~esr/faqs/smart-questions.html
--===============5743251892575364950==--