[jira] [Updated] (TS-4106) Cache Inspector fails to split multiline URLs

["Hiroaki Nakamura (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/TS-4106?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Hiroaki Nakamura updated TS-4106:
---------------------------------
    Attachment: cache_inspector_multiline_urls_split.patch

> Cache Inspector fails to split multiline URLs
> ---------------------------------------------
>
>                 Key: TS-4106
>                 URL: https://issues.apache.org/jira/browse/TS-4106
>             Project: Traffic Server
>          Issue Type: Bug
>          Components: Web UI
>            Reporter: Hiroaki Nakamura
>         Attachments: cache_inspector_multiline_urls_split.patch
>
>
> The error is in the length parameter at
> https://github.com/apache/trafficserver/blob/413dd51d5dc17bf388805071efdb8f882014b847/iocore/cache/CachePages.cc#L129
> Because of this error, ink_strlcpy copies strings from the current poisition to the end of multiline URLs, not to the end of line.
> The patch below fixes this bug:
> {code:title=cache_inspector_multiline_urls_split.patch}
> --- iocore/cache/CachePages.cc.orig     2015-12-31 21:54:56.881997421 +0900
> +++ iocore/cache/CachePages.cc  2015-12-31 21:55:53.574506731 +0900
> @@ -126,7 +126,7 @@
>            q = strstr(p, "%0D%0A" /* \r\n */); // we used this in the JS to separate urls
>            if (!q)
>              q = t;
> -          ink_strlcpy(show_cache_urlstrs[s], p, sizeof(show_cache_urlstrs[s]));
> +          ink_strlcpy(show_cache_urlstrs[s], p, q ? q - p + 1: sizeof(show_cache_urlstrs[s]));
>            p = q + 6; // +6 ==> strlen(%0D%0A)
>          }
>        }
> {code}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)