559469 - Weird results from giving ":" as a hostname

Status: NEW

(Core :: Networking, defect, P3)

Description

[Neil Harris]

While checking out Bug 554596 for edge cases, I thought I'd take a look at IPv6-encapsulated IPv4 addresses. I've come up with this test case:

http://:

Putting it in the Firefox URL bar causes Firefox to rewrite it as 

http://[www.:.com]/

and give the error message "Server not found. Firefox can't find the server at www.:.com." Firefox seems to be confused as to whether ":" is a hostname, or an IPv6 address -- either way, the result is wrong, and indicative of a confused code path that may have other issues.

Comment 1

[Neil Harris]

Note that the "http://:" in the comment above is not clickable: this is because Bugzilla has not recognized the whole thing, autolinking instead only to "http://"

Comment 2

[Boris Zbarsky [:bzbarsky]]

URI fixup just takes the reported host from the original URI (which is ":"), makes it into "www.:.com" and calls nsStandardURL::SetHost.

SetHost calls nsStandardURL::EscapeIPv6, which is:

397     nsStandardURL::EscapeIPv6(const char *host, nsCString &result)
398     {
399         // Escape IPv6 address literal by surrounding it with []'s
400         if (host && (host[0] != '[') && PL_strchr(host, ':')) {
401             result.Assign('[');
402             result.Append(host);
403             result.Append(']');
404             return PR_TRUE;
405         }

This is the only caller of EscapeIPv6, which is why the original URI didn't end up as http://[:].

I suppose EscapeIPv6 could do more vetting, but is there a serious problem with the current behavior?  Or more precisely, is there a problem with putting '[' and ']' around any string passed to SetHost that contains ':'?

Comment 3

[Firefox Bug Husbandry Bot]

Bulk change to priority: https://bugzilla.mozilla.org/show_bug.cgi?id=1399258

Comment 4

[Firefox Bug Husbandry Bot]

Bulk change to priority: https://bugzilla.mozilla.org/show_bug.cgi?id=1399258