Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jra_samba
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
11 ms
·
61.
▲
by
jra_samba
5y ago
Yes. Write test code for glibc. You'll find it always returns one of the two. There is a reason for that.
62.
▲
by
jra_samba
5y ago
No, that's not true. If you know anything about C and writing secure code (which I what I was probing for) you know about using malloc(). You know because you have to know something about the internals of memory management. Imagine you
63.
▲
by
jra_samba
5y ago
Candidate answers to this also tell me if they understand anything about malloc implementations, which is a very useful skill to have as a C coder.
64.
▲
by
jra_samba
5y ago
True, it's implementation defined. But almost all implementations choose one of the two options. There is a reason for this. FYI. I introduced a (bad) security bug into Samba a long time ago by expecting the wrong return (i.e. the one
65.
▲
by
jra_samba
5y ago
Both NULL and a unique pointer value can be safely passed to free() :-). Answers to this question taught me about the candidates taste and understanding of good API design :-). Both NULL and "unique pointer" are correct answers, b
66.
▲
by
jra_samba
5y ago
When I did interviews @ Google (I only do hiring committee work now, thank god :-) I usually asked questions around bugs I added to an existing codebase, to see if the candidate can avoid the pitfalls I ran into by making bad assumptions. A
67.
▲
by
jra_samba
5y ago
I'm not sure it's a new bug, just that if you look up that call you'll find if it's used with any path containing a "/", then it can be raced to check somewhere other that the place the original author int
68.
▲
by
jra_samba
5y ago
That doesn't always fix it. An attacker can race you to make you write something in a place you didn't intend or expect unless the application is incredibly carefully written. And by "incredibly" I mean beyond the scop
69.
▲
by
jra_samba
5y ago
I'm going to be giving a talk at SambaXP this year (it's virtual, so you only need to register to attend) explaining why IMHO symlinks have utterly broken the POSIX filesystem API, making it impossible for application developers t
70.
▲
by
jra_samba
5y ago
is_path_owned_by_current_uid(const char *path) isn't symlink safe given a multi-component path. Symlinks, the poisonous gift that keeps on giving.
71.
▲
by
jra_samba
5y ago
Just wanted to let everyone know about a milestone in coding achieved by David Mulder from SuSE. David reconfigured our codebase to allow the SMB1 code inside smbd to be removed using an option at configure time ! Configure Samba using: con
72.
▲
SMB1 can now be removed from the Samba server code
(lists.samba.org)
6 points
by
jra_samba
5y ago
|
1 comments
73.
▲
by
jra_samba
5y ago
Ugh. An utterly unparsable language designed to insert security-holes into applications using it :-).
74.
▲
by
jra_samba
5y ago
Well for Apples motives you would have to ask Apple. But from an external view of Apple, yes - this does seem to be a direct response to the GPLv3 anti-tivoization clause. Again, this is perfectly within their rights so I don't want an
75.
▲
by
jra_samba
5y ago
Well they didn't have to move, just to be captain pedantic :-). They chose to do so, as they made a policy decision not to use any GPLv3 code. Which is their right to do so, of course. Apple have been moving to only permissively lice
76.
▲
by
jra_samba
5y ago
Microsoft Office (local versions of course) is the main offender here :-). I'm not sure how much testing they do against non-Windows servers.
77.
▲
by
jra_samba
5y ago
Are you using signing/encryption ? Never mind, saw the local "gigabit connection" above, so wired.
78.
▲
by
jra_samba
5y ago
For merely copying files NFS is fine (personally I use scp for that). But when you need to host content that Windows apps want to use with full Windows semantics, then SMB is the way to go.
79.
▲
by
jra_samba
5y ago
More details please. What server are you connecting to locally ?
80.
▲
by
jra_samba
5y ago
Putting it in the vfs objects = list causes the module to be loaded, so enabling the vulnerability.
81.
▲
by
jra_samba
5y ago
No it is not enabled by default.
82.
▲
by
jra_samba
5y ago
Then you are opening yourself up to a world of subtle security holes with symlinks. The POSIX file system API is a festering sore, an utter disaster. Symlinks are the cause.
83.
▲
by
jra_samba
5y ago
I bought 3 OSMC Vero 4k boxes ( https://osmc.tv/vero/ ). Basically a dedicated Kodi box on ARM hardware. Nothing but a happy customer, I use them to play content from my dedicated home server over SMB2.
84.
▲
Performance comparison between Linux in-kernel SMB server and Samba
(samba.plus)
6 points
by
jra_samba
5y ago
|
0 comments
85.
▲
by
jra_samba
5y ago
Prepare to be surprised :-). https://sfconservancy.org/copyleft-compliance/ Oh, I may have misunderstood you there, apologies if so. Conservancy can handle compliance and enforcement activities on behalf of more than a
86.
▲
by
jra_samba
5y ago
Full legal text of the complaint: https://sfconservancy.org/docs/conservancy-v-vizio-original-... Press kit: https://shoestring.agency/wp-content/uploads/2021/10/SFC_Pre...
87.
▲
Software Freedom Conservancy files GPL lawsuit against Vizio
(sfconservancy.org)
347 points
by
jra_samba
5y ago
|
115 comments
88.
▲
by
jra_samba
5y ago
Try using the "legacy" method, just to ensure you can get it working. Using "vfs objects = ceph" causes Samba to load a custom VFS module that translates SMB1/2/3 requests -> Samba VFS -> cephfs calls. Th
89.
▲
by
jra_samba
5y ago
"Dave, this conversation can serve no purpose anymore. Goodbye."
90.
▲
by
jra_samba
5y ago
Samba ships with modules for ceph, gluster and gpfs. Here's the info you need: https://documentation.suse.com/ses/6/html/ses-all/cha-ses-ci... Note the share definition: [SHARE_NAME] path = /
More ›