[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH 2/8] block: fix unbounded stack for dump_qdict
From: |
Paolo Bonzini |
Subject: |
Re: [Qemu-devel] [PATCH 2/8] block: fix unbounded stack for dump_qdict |
Date: |
Tue, 8 Mar 2016 13:17:03 +0100 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.6.0 |
On 08/03/2016 09:12, Markus Armbruster wrote:
> I'm afraid this isn't a good idea. It relies on the non-local argument
> that nobody will ever put a key longer than 255 into a qdict that gets
> dumped. That may even be the case, but you need to *prove* it, not just
> assert it. The weakest acceptable proof might be assertions in every
> place that put keys into a dict that might get dumped. I suspect that's
> practical and maintainable only if there's a single place that does it.
>
> If this was a good idea, I'd recommend to avoid the awkward macro:
>
> char key[256];
> int i;
>
> assert(strlen(entry->key) + 1 <= ARRAY_SIZE(key));
>
> There are several other ways to limit the stack usage:
>
> 1. Move the array from stack to heap. Fine unless it's on a hot path.
> As far as I can tell, this dumping business is for HMP and qemu-io,
> i.e. not hot.
I think this is the best. You can just g_strdup, modify in place, print
and free.
Paolo
- Re: [Qemu-devel] [PATCH 1/8] qdict: fix unbounded stack for qdict_array_entries, (continued)
[Qemu-devel] [PATCH 2/8] block: fix unbounded stack for dump_qdict, Peter Xu, 2016/03/08
Re: [Qemu-devel] [PATCH 2/8] block: fix unbounded stack for dump_qdict, Peter Xu, 2016/03/08
Re: [Qemu-devel] [PATCH 2/8] block: fix unbounded stack for dump_qdict, Markus Armbruster, 2016/03/08
Re: [Qemu-devel] [PATCH 2/8] block: fix unbounded stack for dump_qdict,
Paolo Bonzini <=
Re: [Qemu-devel] [PATCH 2/8] block: fix unbounded stack for dump_qdict, Peter Xu, 2016/03/08
[Qemu-devel] [PATCH 3/8] usb: fix unbounded stack for ohci_td_pkt, Peter Xu, 2016/03/08
[Qemu-devel] [PATCH 4/8] usb: fix unbounded stack for xhci_dma_write_u32s, Peter Xu, 2016/03/08