Robert Elz wrote:
> Date:Tue, 8 May 2018 19:15:28 +0100
> From:Alexander Nasonov
> Message-ID: <20180508180815.GA5990@neva>
>
> | I think it broke the tool. If you run
> |
> | cgdconfig -p
> |
> | it will crash.
>
> Sorry, I cannot reproduce this, it looks
Alexander Nasonov wrote:
> (gdb) b opendisk1
> (gdb) run -p
> Starting program:
> /home/alnsn/netbsd-current/clean/src/sbin/cgdconfig/obj/cgdconfig -p
>
> Breakpoint 1, 0x7f7ff78111f6 in opendisk1 () from /lib/libutil.so.7
> (gdb) x/s $rdi
> 0x0: # path=NULL
Adding (argc > 0) check be
Date:Wed, 9 May 2018 08:59:55 +0100
From:Alexander Nasonov
Message-ID: <20180509075955.GA7743@neva>
| Adding (argc > 0) check before calling opendisk1 fixes the crash.
Thanks - and I see what is wrong now, but (for whatever reason) that did
not fail for me, I guess
"Alexander Nasonov" writes:
> Module Name: src
> Committed By: alnsn
> Date: Wed May 9 18:11:56 UTC 2018
>
> Modified Files:
> src/sbin/cgdconfig: cgdconfig.8 cgdconfig.c
>
> Log Message:
> Add '-e' option (echo the passphrase) and wipe the passphrase after use.
>
> XXX Using me
matthew green wrote:
> "Alexander Nasonov" writes:
> > XXX Using memset for wiping isn't a good idea because memset is likely
> > optimised away by gcc. This should be revisited.
>
> use explicit_memset(3)?
Yes, we should change memsets of sensitive buffers to explicit_memset
but we also should i
