Messages in this thread | | | Date | Thu, 01 Feb 2018 20:32:05 -0600 | From | "Gustavo A. R. Silva" <> | Subject | Re: [PATCH] tcp_lp: use 64-bit arithmetic instead of 32-bit |
| |
Hi Andrew,
Quoting Andrew Lunn <andrew@lunn.ch>:
> On Wed, Jan 31, 2018 at 07:07:49PM -0600, Gustavo A. R. Silva wrote: >> >> Hi Alan, >> >> Quoting Alan Cox <gnomes@lxorguk.ukuu.org.uk>: >> >> >On Wed, 31 Jan 2018 18:24:07 -0600 >> >"Gustavo A. R. Silva" <gustavo@embeddedor.com> wrote: >> > >> >>Cast to s64 some variables and a macro in order to give the >> >>compiler complete information about the proper arithmetic to >> >>use. Notice that these elements are used in contexts that >> >>expect expressions of type s64 (64 bits, signed). >> >> >> >>Currently such expression are being evaluated using 32-bit >> >>arithmetic. >> > >> >The question you need to ask is 'can it overflow 32bit maths', otherwise >> >you are potentially making the system do extra work for no reason. >> > >> >> Yeah, I get your point and it seems that in this particular case there is no >> risk of a 32bit overflow, but in general and IMHO as the code evolves, the >> use of incorrect arithmetic may have security implications in the future, so >> I advocate for code correctness in this case. > > Hi Gustavo > > Is this on the hotpath? How much overhead does it add to 32 bit > architectures which don't have 64 bit arithmetic in hardware? There > are a lot of embedded systems which are 32 bit. >
I'm sorry, I don't have access to 32-bit hardware at the moment.
Thanks -- Gustavo
| |