-
Notifications
You must be signed in to change notification settings - Fork 247
ACL fixes #465
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ACL fixes #465
Conversation
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lls: fix ICMP/ICMPv6 ACL registration
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
acl/net: be careful about ntuple filters and ACLs
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
acl/net: coalesce ACL/ntuple into one abstraction
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Patch acl: update rule registration interface is ready for merge.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
acl/net: coalesce ACL/ntuple into one abstraction
-
With the exception of
lib/net.cand the GK and GT blocks, we should be able to stop including the headergatekeeper_acl.heverywhere else. -
The LLS block is not dealing with ICMP packets that may arrive now that it registers for ntuple filters. Given the current patch is already large, it would be better to just add a
TODOcomment in this patch, and address this issue in a following patch.
|
I removed |
fa11ff2 to
e4c5e17
Compare
|
Ready for another review (but is not yet tested). |
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
acl/net: be careful about ntuple filters and ACLs
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
acl/net: coalesce ACL/ntuple into one abstraction
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
acl/net: coalesce ACL/ntuple into one abstraction
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
acl/net: coalesce ACL/ntuple into one abstraction
fe63570 to
135db9b
Compare
|
Ready for another review. Tested on Amazon. |
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
net: define RX methods for blocks
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
net: combine ntuple filters/ACL into packet filters
We should be able to remove gatekeeper_acl.h from the GGU this time around.
It's worth verifying that only lib/net.c and the GK and GT blocks are adding the header gatekeeper_acl.h.
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lls: use packet filter API
38e67b6 to
6ad2d86
Compare
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
net: define RX methods for blocks
|
Shouldn't we remove line |
|
My fault, it has already been removed. |
AltraMayor
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
net: combine ntuple filters/ACL into packet filters
|
Please fix the last comment, and run a test on Amazon. |
In at least some current NICs (Intel 82599) the IPv6 destination IP field cannot be matched by ntuple filters (AltraMayor#466). Therefore, it is possible for ntuple filters to appear to be available but can not be used, e.g., in an IPv6-only deployment. This patch makes the code more careful about this distinction, and defines RX methods so that blocks can decide whether they need to query NICs or mailboxes, for example, to receive packets.
Since ntuple filters can be mostly equivalent to ACL rules when hardware does not support ntuple filters, this patch coalesces their creation into a single API for packet filters, including one function for IPv4 and one for IPv6, so that blocks don't have to worry about what hardware supports. Closes AltraMayor#63.
|
Fixed and tested. Ready for another review. |
No description provided.