| Summary: | SOCKS4A support | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | Portable OpenSSH | Reporter: | David Woodhouse <dwmw2> | ||||||
| Component: | ssh | Assignee: | Assigned to nobody <unassigned-bugs> | ||||||
| Status: | CLOSED FIXED | ||||||||
| Severity: | enhancement | CC: | djm | ||||||
| Priority: | P2 | ||||||||
| Version: | 5.0p1 | ||||||||
| Hardware: | Other | ||||||||
| OS: | Linux | ||||||||
| Bug Depends on: | |||||||||
| Bug Blocks: | 1481 | ||||||||
| Attachments: |
|
||||||||
Created attachment 1544 [details]
correct patch
Oops; wrong patch. Try this one. Sorry.
patch applied - this will be in openssh-5.2 Close bugs fixed/reviewed for openssh-5.2 release |
Created attachment 1543 [details] patch This provides SOCKS4A support, which allows a SOCKS4 client to provide a host name instead of an IP address. It indicates that it's doing so by using a known-invalid IP address 0.0.0.x (where x != 0), and then appending the NUL-terminated hostname to the connection request. Although we check for the terminating NUL manually and the 'fatal(len %d > have %d)' should never happen, the condition for it (in the existing processing of the username) seems to have been off-by-one; not accounting for the terminating NUL. I've fixed that while I'm at it.