Re: RISC-V Vector Extension post-public review updates

Krste Asanovic

I guess simpler examples are anytime you use v0 as mask and a data

These aren't useful use-cases, so existing software shouldn't have
been doing this (except test code).


On Mon, 15 Nov 2021 15:45:14 -0800, Nick Knight <nick.knight@...> said:
| On Mon, Nov 15, 2021 at 3:40 PM Krste Asanovic <krste@...> wrote:
| I'm not sure if C intrinsics can generate this case,


| but there are
| other cases where dynamic value settings can result in illegal
| instruction traps, so the result would be the same that
| implementations will either trap or do something non-conforming.

| Krste

|||||| On Mon, 15 Nov 2021 15:28:25 -0800, Craig Topper <craig.topper@...> said:

| | On Nov 15, 2021, at 3:24 PM, Krste Asanovic <krste@...> wrote:
| |         On Nov 15, 2021, at 3:13 PM, Guy Lemieux <guy.lemieux@...> wrote:

| |             On Mon, Nov 15, 2021 at 2:17 PM Bill Huffman <huffman@...> wrote:

| |                 I'm glad this came up.  I certainly wouldn't want to try to make an implementation work for these cases.  😊

| |                 I lean a bit toward #3, not so much because we might use the space as because I think we've called all the other similar
| |                 corners of opcode space that don't make sense to implement "reserved."  Possibly that's because they might make sense
| someday
| |                 and this won't.

| |             I think these encodings are qualitatively different from other nooks and crannies, since their availability is a function of
| the
| |             dynamic vtype setting.  So we can rationalize the departure from the normal practice of marking the state reserved.

| |         Ok, this makes the opcodes virtually useless for other instructions.

| |         Instead, shouldn't we be setting a bit similar to vill?  I realize vill is only set on illegal vset* instructions; in this case
| it
| |         would be a new bit which is only set on executing instructions that are incompatible with the current (but otherwise valid) vtype
| ?

| |         Guy

| |     There’s no benefit to setting vill versus just taking a trap in this case.

| |     Vill is there so we don’t have to add the first trap on a write of a particular data value, and also to provide a discovery
| mechanism.

| |     Krste

| | Is it possible to generate one of these cases from C with crazy uses of vreinterpret and vget/vset intrinsics? What should the compiler
| do for
| | such code?

| | Craig

| |     


Join to automatically receive all group messages.