Thread (63 messages) 63 messages, 7 authors, 2025-09-09

Re: [PATCH v6 5/6] tracing: Show inode and device major:minor in deferred user space stacktrace

From: Steven Rostedt <rostedt@goodmis.org>
Date: 2025-08-30 01:25:49
Also in: bpf, lkml

On Fri, 29 Aug 2025 21:20:23 -0400
Steven Rostedt [off-list ref] wrote:
quoted
I'm done arguing. You're not listening, so I'll just let you know that  
I am listening. I'm just not understanding you.
BTW, I'm not arguing with you. I'm really trying hard to figure out what it
is that you want me to do. I'm looking for that "don't use dentry outside
of VFS" moment.

I get we have in the call stack the offsets of the file and a magical hash
value that represents that vma.

What I don't get is what is user space suppose to match that magical hash
value to?

Do you want me to trace all mmaps and trigger an event for them that show
the hash value and the path names?

If that's the case, what do I do about the major use case of tracing an
application after it has mapped all it's memory to files?

What about wasted ring buffer space for recording every mmap when the
majority of them will not be used. It could risk dropping events of the
mmaps we care about.

Again, I'm not arguing with you. I'm trying to figure out what you are
suggesting.

-- Steve
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help