Re: [PATCH v6 11/11] arm64: annotate user pointers casts detected by sparse

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, Sep 06, 2018 at 03:13:16PM +0100, Vincenzo Frascino wrote:
> On 05/09/18 20:03, Luc Van Oostenryck wrote:
> > On Tue, Sep 04, 2018 at 12:27:23PM +0100, Vincenzo Frascino wrote:
> >> ---
> >> commit ee7985f0c2b29c96aefe78df4139209eb4e719d8
> >> Author: Vincenzo Frascino <vincenzo.frascino@xxxxxxx>
> >> Date:   Wed Aug 15 10:55:44 2018 +0100
> >>
> >>     print address space number for explicit cast to ulong
> >>     
> >>     This patch build on top of commit b34880d ("stricter warning
> >>     for explicit cast to ulong") and prints the address space
> >>     number when a "warning: cast removes address space of expression"
> >>     is triggered.
> >>     
> >>     This makes easier to discriminate in between different address
> >>     spaces.
> >>     
> >>     A validation example is provided as well as part of this patch.
> >>     
> >>     Signed-off-by: Vincenzo Frascino <vincenzo.frascino@xxxxxxx>
> >>
> >> diff --git a/evaluate.c b/evaluate.c
> >> index 6d5d479..2fc0ebc 100644
> >> --- a/evaluate.c
> >> +++ b/evaluate.c
> >> @@ -3017,8 +3017,12 @@ static struct symbol *evaluate_cast(struct expression *expr)
> >>  		sas = stype->ctype.as;
> >>  	}
> >>  
> >> -	if (!tas && sas > 0)
> >> -		warning(expr->pos, "cast removes address space of expression");
> >> +	if (!tas && sas > 0) {
> >> +		if (Wcast_from_as)
> >> +			warning(expr->pos, "cast removes address space of expression (<asn:%d>)", sas);
> >> +		else
> >> +			warning(expr->pos, "cast removes address space of expression");
> >> +	}
> > 
> > I think that the if (Wcast_from_as) is unneeded, the <asn:%d> can be added
> > even if Wcast_from_as is false. Would it be OK for you?
> > 
> 
> Yes, it is OK for me (I put it there because I did not know if we wanted to preserve the
> original behavior). Feel free to hack my patch if you want to put it on your tree.
> Thanks.

Yes, adding the asn is a good thing and the content of the original message
is even not changed, just extended. I have since long the plan to associate
a name to the address space and then to display this name when one is defined
but this will need to wait.

Thanks,
-- Luc



[Index of Archives]     [Newbies FAQ]     [LKML]     [IETF Annouce]     [DCCP]     [Netdev]     [Networking]     [Security]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [Trinity Fuzzer Tool]

  Powered by Linux