From: Eric Bollengier <eric@baculasystems.com>
Date: Wed, 18 Jan 2012 09:31:32 +0000 (+0100)
Subject: Revert to 5.0 size computation in accurate elements, may fix #1821
X-Git-Url: https://git.sur5r.net/?a=commitdiff_plain;h=0aab02e7978a7a2a83580fb8d0c8f05013d05504;p=bacula%2Fbacula

Revert to 5.0 size computation in accurate elements, may fix #1821
---

diff --git a/bacula/src/filed/accurate.c b/bacula/src/filed/accurate.c
index 8bdc7a6c51..6a2c9e6738 100644
--- a/bacula/src/filed/accurate.c
+++ b/bacula/src/filed/accurate.c
@@ -215,7 +215,7 @@ static bool accurate_add_file(JCR *jcr, uint32_t len,
    /* we store CurFile, fname and ctime/mtime in the same chunk 
     * we need one extra byte to handle an empty chksum
     */
-   item = (CurFile *)jcr->file_list->hash_malloc(sizeof(CurFile)+len+1);
+   item = (CurFile *)jcr->file_list->hash_malloc(sizeof(CurFile)+len+3);
    item->seen = 0;
 
    /* TODO: see if we can optimize this part with memcpy instead of strcpy */