This looks good enough. Is this related to the recently reported crash bug? Is there a test somewhere that can trigger an out of bounds read or write?
On Oct 8, 1:32 pm, Mukund Sivaraman <[EMAIL PROTECTED]> wrote: > --- > server/items.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/server/items.c b/server/items.c > index 88c92f6..42f7416 100644 > --- a/server/items.c > +++ b/server/items.c > @@ -303,7 +303,7 @@ char *do_item_cachedump(const unsigned int slabs_clsid, > const unsigned int limit > unsigned int shown = 0; > char temp[512]; > > - if (slabs_clsid > LARGEST_ID) return NULL; > + if (slabs_clsid >= LARGEST_ID) return NULL; > it = heads[slabs_clsid]; > > buffer = malloc((size_t)memlimit); > -- > 1.5.5.1
