<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
</head>
<body style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;" class="">
<div class="">Oops, vscale * n * sizeof(ty).</div>
<div class=""><br class="">
</div>
<div class="">:)</div>
<br class="">
<div><br class="">
<blockquote type="cite" class="">
<div class="">On 12 Jun 2018, at 18:18, Graham Hunter <<a href="mailto:Graham.Hunter@arm.com" class="">Graham.Hunter@arm.com</a>> wrote:</div>
<br class="Apple-interchange-newline">
<div class="">
<div style="word-wrap: break-word; -webkit-nbsp-mode: space; line-break: after-white-space;" class="">
<div class="">Hi,</div>
<div class=""><br class="">
</div>
<div class="">Just to clarify the part below about changing the proposal, this means that a <scalable n x ty> vector:</div>
<div class=""><br class="">
</div>
<div class="">* Has a size of vscale * sizeof(ty) when reasoning about sizes within llvm</div>
<div class="">* Potentially has a smaller size of active_vscale * sizeof(ty) depending on the current</div>
<div class=""> processor state at runtime.</div>
<div class=""><br class="">
</div>
<div class="">For SVE, active_vscale == vscale. For RVV (and potentially others), it can be smaller based</div>
<div class="">on the VL state generated by setvl or similar.</div>
<div class=""><br class="">
</div>
<div class="">-Graham</div>
<div class=""><br class="">
</div>
<br class="">
<div class="">
<blockquote type="cite" class="">
<div class="">
<blockquote type="cite" style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-size-adjust: auto; -webkit-text-stroke-width: 0px;" class="">
I am not quite so sure about turning the active vector length into<br class="">
just another mask. It's true that the effects on arithmetic, load,<br class="">
stores, etc. are the same as if everything executed under a mask like<br class="">
<1, 1, ..., 1, 0, 0, ..., 0> with the number of ones equal to the<br class="">
active vector length. However, actually materializing the masks in the<br class="">
IR means the RISCV backend has to reverse-engineer what it must do<br class="">
with the vl register for any given (masked or unmasked) vector<br class="">
operation. The stakes for that are rather high, because (1) it applies<br class="">
to pretty much every single vector operation ever, and (2) when it<br class="">
fails, the codegen impact is incredibly bad.<br class="">
</blockquote>
<br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">I
can see where the concern comes from; we had problems reconstructing</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">semantics
when experimenting with search loop vectorization and often</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">had
to fall back on default (slow) generic cases.</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">My
main reason for proposing this was to try and ensure that the size was</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">consistent
from the point of view of the query functions we were discussing</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">in
the main thread. If you're fine with all size queries assuming maxvl (so</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">things
like stack slots would always use the current configured maximum</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">length),
then I don't think there's a problem with dropping this part of the</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
<span style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">proposal
and letting you find a better representation of active length.</span><br style="font-family: Menlo-Regular; font-size: 16px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class="">
</div>
</blockquote>
</div>
<br class="">
</div>
</div>
</blockquote>
</div>
<br class="">
</body>
</html>