Date: Mon, 8 Jun 2009 11:31:27 +0400 From: Eygene Ryabinkin <rea-fbsd@codelabs.ru> To: hrs@freebsd.org Cc: amd64@freebsd.org, current@freebsd.org Subject: Re: [head tinderbox] failure on amd64/amd64 Message-ID: <sqWp8ZLqcJhvIOZAWr3MfqIXJ9k@XX1fo6zQUfC4h0jjRC6IBz3oNH4> In-Reply-To: <8LPG99US2/4EsGlonyfMSkDb40o@XX1fo6zQUfC4h0jjRC6IBz3oNH4> References: <20090608025715.499087302F@freebsd-current.sentex.ca> <8LPG99US2/4EsGlonyfMSkDb40o@XX1fo6zQUfC4h0jjRC6IBz3oNH4>
next in thread | previous in thread | raw e-mail | index | archive | help
Mon, Jun 08, 2009 at 11:00:56AM +0400, Eygene Ryabinkin wrote:
> Looks like that
> -----
> ifr->ifr_data = (caddr_t)(size_t)options;
> -----
> will be more correct and will disable this warning -- it will convert
> u_int to the proper type that will be able to carry addresses for the
> given platform.
Hmm, looking a bit into the code of gif_ioctl, I am under impression
that 'options' will not be initialized at the GIFSOPTS processing.
And the statement
-----
if ((error = copyin(&options, &sc->gif_options,
sizeof(sc->gif_options)))) {
-----
looks strange -- (&options) is in the kernel space (stack space),
so why one is passing it as the userland address? Judging by the
contents of newly added setgifopts() inside ifgif.c, I would assume
that one wants 'ifr->ifr_data' instead of '&options'.
Am I missing something?
--
Eygene
_ ___ _.--. #
\`.|\..----...-'` `-._.-'_.-'` # Remember that it is hard
/ ' ` , __.--' # to read the on-line manual
)/' _/ \ `-_, / # while single-stepping the kernel.
`-'" `"\_ ,_.-;_.-\_ ', fsc/as #
_.-'_./ {_.' ; / # -- FreeBSD Developers handbook
{_.-``-' {_/ #
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?sqWp8ZLqcJhvIOZAWr3MfqIXJ9k>