Re: [RFC] arm64: Detecting tagged addresses

2019-06-29 Thread Vincenzo Frascino
+ Dan and sma...@vger.kernel.org Hi Andrew, I am adding Dan to this thread since he is the smatch maintainer, and the sma...@vger.kernel.org list. @Dan and @sma...@vger.kernel.org: a reference to the beginning of this thread can be found at [1]. [1] https://lkml.org/lkml/2019/6/19/376 On

Re: [RFC] arm64: Detecting tagged addresses

2019-06-27 Thread Andrew Murray
On Wed, Jun 26, 2019 at 06:45:03PM +0100, Catalin Marinas wrote: > Hi Andrew, > > Cc'ing Luc (sparse maintainer) who's been involved in the past > discussions around static checking of user pointers: > > https://lore.kernel.org/linux-arm-kernel/20180905190316.a34yycthgbamx2t3@ltop.local/ > > So

Re: [RFC] arm64: Detecting tagged addresses

2019-06-26 Thread Catalin Marinas
Hi Andrew, Cc'ing Luc (sparse maintainer) who's been involved in the past discussions around static checking of user pointers: https://lore.kernel.org/linux-arm-kernel/20180905190316.a34yycthgbamx2t3@ltop.local/ So I think the difference here from the previous approach is that we explicitly

[RFC] arm64: Detecting tagged addresses

2019-06-19 Thread Andrew Murray
Hello, The proposed introduction of a relaxed ARM64 ABI [1] will allow tagged memory addresses to be passed through the user-kernel syscall ABI boundary. Tagged memory addresses are those which contain a non-zero top byte (the hardware has always ignored this top byte due to TCR_EL1.TBI0) and may