<div style="line-height:1.7;color:#000000;font-size:14px;font-family:Arial"><div style="margin:0;">Hello,</div><div style="margin:0;"><br></div><div style="margin:0;">Though the ABI / call-convention for rv32e has not be defined clear by the riscv organization. </div><div style="margin:0;"><br></div><div style="margin:0;">But besides that, is it possible to partial support it by adding target feature FeatureReserveX16---FeatureReserveX31 to the FeatureRV32E ?</div><div style="margin:0;">Just like </div><div style="margin:0;"><div style="margin:0;">def FeatureStdExtD</div><div style="margin:0;">    : SubtargetFeature<"d", "HasStdExtD", "true",</div><div style="margin:0;">                       "'D' (Double-Precision Floating-Point)",</div><div style="margin:0;">                       [FeatureStdExtF]>;</div><div style="margin:0;"><br></div><div style="margin:0;">FeatureStdExtD implies FeatureStdExtF, we can make FeatureRV32E implies FeatureReserveX16---FeatureReserveX31 ? </div><div style="margin:0;"><br></div><div style="margin:0;">Is this a proper way to tell the register allocator do not use X16-X31 ?</div><div style="margin:0;"><br></div><div style="margin:0;">Ben</div><div style="margin:0;"><br></div></div></div>