<div><div dir="auto">Hi Chandler,</div><div dir="auto">I'm just wondering if you have any updates regarding this patch. Have you had a chance to try it out?</div><div dir="auto"><br></div><div dir="auto">Nemanja </div><br><div class="gmail_quote"><div>On Tue, Aug 15, 2017 at 9:22 PM Nemanja Ivanovic <<a href="mailto:nemanja.i.ibm@gmail.com">nemanja.i.ibm@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div><div><div><div><div>Hi Chandler,<br></div>here's a patch that includes the changeset you reverted as well as the fix. I would really appreciate it if you have time to test it out with the library in which you had miscompiles, to see if this fixes the issue.<br></div>Once again, I am really sorry about the work this generated for you.<br><br></div>Nemanja<br><br></div>P.S. It is very difficult to see why changing the operand for RLDICL to 33 doesn't work in every case, but here's a test case that illustrates that:<br><span style="font-family:monospace,monospace">int test(int a, int b) {<br> return (a << 24) <= (b << 24);<br>}</span><br></div>Calling this function with arguments -1, 127 will return zero instead of 1.<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Aug 15, 2017 at 12:48 AM, Nemanja Ivanovic <span><<a href="mailto:nemanja.i.ibm@gmail.com" target="_blank">nemanja.i.ibm@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div dir="auto">I can certainly whip together the original patch with the fix tomorrow. I'll send you that patch tomorrow. </div><div dir="auto"><br></div><div dir="auto">And please observe that R5 is reloaded (lwz - zero-extending load) and it's an operand to subf. I assume that if you were to track down where that was spilled, the stw would be preceded by an extsw (or it was otherwise sign-extended, such as coming from a parameter).</div><span class="m_-7613560513523954450HOEnZb"><font color="#888888"><div dir="auto"><br></div><div dir="auto">Nemanja</div></font></span><div><div class="m_-7613560513523954450h5"><br><div class="gmail_quote"><div>On Tue, Aug 15, 2017 at 12:37 AM Chandler Carruth <<a href="mailto:chandlerc@gmail.com" target="_blank">chandlerc@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div class="gmail_quote"><div>On Mon, Aug 14, 2017 at 3:03 PM Nemanja Ivanovic <<a href="mailto:nemanja.i.ibm@gmail.com" target="_blank">nemanja.i.ibm@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div><div><div><div><div><div>Hi Chandler,<br></div>thanks for investigating this.<br><br></div>The 32-bit values should be sign-extended correctly in the register and a negative value should contain a 1 in bit 0 (the high bit in a 64-bit register).</div></div></div></div></div></blockquote><div><br></div></div></div><div><div class="gmail_quote"><div>I'm not a PPC expert, so I believe you, but this doesn't seem to match my experimental data...</div></div></div><div><div class="gmail_quote"><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div><div><div><div> I believe that the likely reason for the failures you're seeing is the issue I only discovered late last week (and have a fix for in patch <a href="https://reviews.llvm.org/D36613" target="_blank">https://reviews.llvm.org/D36613</a>). Namely, the issue is that we didn't change the register class when sign-extending the values so they may be spilled somewhere mid-way through this sequence. That has the effect of zero-extending the value at the point of the restore.<br></div></div></div></div></div></blockquote><div><br></div></div></div><div><div class="gmail_quote"><div>I don't think this is the issue. Here is the before/after assembly diff from my test case (sadly not reduced):</div><div><br></div><div><font face="monospace">```</font></div><div><div><font face="monospace">@@ -39860,13 +39860,17 @@</font></div><div><font face="monospace"> b .LBB24_9</font></div><div><font face="monospace"> .p2align 4</font></div><div><font face="monospace"> .LBB24_6: # in Loop: Header=BB24_3 Depth=2</font></div><div><font face="monospace">- lwz 15, 668(31) # 4-byte Folded Reload</font></div><div><font face="monospace">- clrlwi 3, 30, 24</font></div><div><font face="monospace">- subfic 4, 24, -128</font></div><div><font face="monospace">- cmpwi 3, 0</font></div><div><font face="monospace">- cmpw 1, 4, 15</font></div><div><font face="monospace">- crorc 20, 2, 5</font></div><div><font face="monospace">- bc 12, 20, .LBB24_8</font></div><div><font face="monospace">+ subfic 3, 24, -128</font></div><div><font face="monospace">+ lwz 5, 668(31) # 4-byte Folded Reload</font></div><div><font face="monospace">+ clrlwi 4, 30, 24</font></div><div><font face="monospace">+ extsw 3, 3</font></div><div><font face="monospace">+ cntlzw 4, 4</font></div><div><font face="monospace">+ subf 3, 3, 5</font></div><div><font face="monospace">+ srwi 4, 4, 5</font></div><div><font face="monospace">+ rldicl 3, 3, 1, 63</font></div><div><font face="monospace">+ xori 3, 3, 1</font></div><div><font face="monospace">+ or. 3, 4, 3</font></div><div><font face="monospace">+ bc 12, 1, .LBB24_8</font></div><div><font face="monospace"> # BB#7: # in Loop: Header=BB24_3 Depth=2</font></div><div><font face="monospace"> ori 4, 16, 2</font></div><div><font face="monospace"> li 3, 128</font></div></div><div><font face="monospace">```</font></div><div><br></div><div>To me, it certainly looks like `subf` and `rldicl` are not separated by a spill. Changing the `1` operand to `rldicl` to `33` caused this to pass all of the tests of this code...</div><div><br></div><div>So I can't say what the actual issue is here, or if my PPC assembly reading is just inadequate... But it doesn't seem to involve spills.</div><div><br></div><div>Looping in a person who understand PPC much better than I do and has access to this test case.</div></div></div><div><div class="gmail_quote"><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div><div><div><div>Do you think that you could see if that patch fixes the issue you were seeing?<br></div></div></div></div></div></blockquote><div><br></div></div></div><div><div class="gmail_quote"><div>Do you have a new patch that is the combination of the original patch and the fix? That would be the easiest to try.</div></div></div><div><div class="gmail_quote"><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div><div><div><div></div><div><br></div>I am sorry that I didn't pull the changeset immediately upon realizing this issue (or at least opened a bug for it).<br><br></div>A test case that should reproduce the problem:<br><br><span style="font-family:monospace,monospace">int printf(const char *, ...);<br>void __attribute__((noinline)) call(int a) { asm ("#Do Nothing" : : : "memory"); }<br>int __attribute__((noinline)) test(int a, int b, int c) {<br> int Ret = a + b;<br> call(Ret);<br> asm ("#Do Nothing"<br> :<br> :<br> : "r0", "r3", "r4", "r5",<br> "r6", "r7", "r8", "r9",<br> "r10", "r11", "r12", "r13",<br> "r14", "r15", "r16", "r17",<br> "r18", "r19", "r20", "r21",<br> "r22", "r23", "r24", "r25",<br> "r26", "r27", "r28", "r29",<br> "r30", "r31");<br> call(Ret - c);<br> return Ret <= c;<br>}<br><br>int main(void) {<br> int a = 10;<br> int b = -15;<br> int c = 0;<br> printf("Value = %d\n", test(a, b, c));<br> return 0;<br>}</span><br><br></div>It is so contrived so as to ensure that the spill happens in a place where it will cause the failure.<br><br></div>Finally, I understand that viewed in isolation, it seems like a weird choice to emit this code in C++ rather than providing a DAG combine + any patterns in .td files. However, this is a part of a series of patches that produces code for integer comparisons in GPRs rather than CRs. A number of those patterns use instructions that set the CARRY bit (i.e. produce more than one value) and we cannot represent those patterns in .td files. So in order to keep this code all in one place, we implemented it in ISELDAGToDAG.cpp.<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Aug 14, 2017 at 5:43 AM, Chandler Carruth <span><<a href="mailto:chandlerc@gmail.com" target="_blank">chandlerc@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div>Reverted in r310809.<div><br></div><div>Another note about this patch -- why not handle this with a 'Pat' in the backend? It seems like it would be easy to expand the SETLE and SETGE nodes into this pattern there. Similarly, it would seem better to expand into more generic ISDs and let normal instruction selection handle them. For example, I wonder about cases where the extract of the sign bit can be combined with other surrounding logic. This generates a lot of instructions all to avoid the SETLE/SETGE generic expansion. But I'm not a PPC expert, so it's quite likely I just don't understand the implications. =]</div></div><div class="m_-7613560513523954450m_4589660210395583478m_-5635927142143294950m_-3595768300941551850HOEnZb"><div class="m_-7613560513523954450m_4589660210395583478m_-5635927142143294950m_-3595768300941551850h5"><br><div class="gmail_quote"><div>On Sun, Aug 13, 2017 at 4:36 PM Chandler Carruth <<a href="mailto:chandlerc@gmail.com" target="_blank">chandlerc@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div class="gmail_quote"><div>This patch appears to be very deeply and fundamentally broken. See below:</div></div></div><div><div class="gmail_quote"><div><br></div><div>On Tue, Aug 8, 2017 at 4:21 AM Nemanja Ivanovic via llvm-commits <<a href="mailto:llvm-commits@lists.llvm.org" target="_blank">llvm-commits@lists.llvm.org</a>> wrote:<br></div></div></div><div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Modified: llvm/trunk/lib/Target/PowerPC/PPCISelDAGToDAG.cpp<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCISelDAGToDAG.cpp?rev=310346&r1=310345&r2=310346&view=diff" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/PowerPC/PPCISelDAGToDAG.cpp?rev=310346&r1=310345&r2=310346&view=diff</a><br>
==============================================================================<br>
--- llvm/trunk/lib/Target/PowerPC/PPCISelDAGToDAG.cpp (original)<br>
+++ llvm/trunk/lib/Target/PowerPC/PPCISelDAGToDAG.cpp Tue Aug 8 04:20:44 2017<br></blockquote></div></div><div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">@@ -2831,6 +2903,32 @@ SDValue PPCDAGToDAGISel::get32BitZExtCom<br>
return SDValue(CurDAG->getMachineNode(PPC::XORI, dl, MVT::i32, Shift,<br>
getI32Imm(1, dl)), 0);<br>
}<br>
+ case ISD::SETGE: {<br>
+ // (zext (setcc %a, %b, setge)) -> (xor (lshr (sub %a, %b), 63), 1)<br>
+ // (zext (setcc %a, 0, setge)) -> (lshr (~ %a), 31)<br>
+ if(IsRHSZero)<br>
+ return getCompoundZeroComparisonInGPR(LHS, dl, ZeroCompare::GEZExt);<br>
+<br>
+ // Not a special case (i.e. RHS == 0). Handle (%a >= %b) as (%b <= %a)<br>
+ // by swapping inputs and falling through.<br>
+ std::swap(LHS, RHS);<br>
+ ConstantSDNode *RHSConst = dyn_cast<ConstantSDNode>(RHS);<br>
+ IsRHSZero = RHSConst && RHSConst->isNullValue();<br>
+ LLVM_FALLTHROUGH;<br>
+ }<br>
+ case ISD::SETLE: {<br>
+ // (zext (setcc %a, %b, setle)) -> (xor (lshr (sub %b, %a), 63), 1)<br>
+ // (zext (setcc %a, 0, setle)) -> (xor (lshr (- %a), 63), 1)<br>
+ if(IsRHSZero)<br>
+ return getCompoundZeroComparisonInGPR(LHS, dl, ZeroCompare::LEZExt);<br>
+ SDValue Sub =<br>
+ SDValue(CurDAG->getMachineNode(PPC::SUBF, dl, MVT::i32, LHS, RHS), 0);<br>
+ SDValue Shift =<br>
+ SDValue(CurDAG->getMachineNode(PPC::RLDICL_32, dl, MVT::i32, Sub,<br>
+ getI64Imm(1, dl), getI64Imm(63, dl)), 0);<br></blockquote></div></div><div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"></blockquote><div><br></div><div>So, even when operating on a 32-bit values, the RLDICL instruction is fundamentally a 64-bit operation. It is rotating left a 64-bit quantity and then clearing the left 63 bits.</div><div><br></div><div>The code is using it to extract the high (sign) bit to determine whether the subtract produced a negative number. But the high (sign) bit is in bit position 31 not bit position 63. And so rotating left by one bit as this does will reliably (AFAICT) produce a zero in bit position 0 and will clear the bits to the left.</div><div><br></div><div>In other words, this always produces '0', which means that the 'setle' always compiles to "true". =[</div><div><br></div><div>I have a high-level library implementing saturating arithmetic and it manages to trigger this reliably. What is worse, it is *extraordinarily* hard to trigger this particular combine from high level code. I tried for a long time (about 8 hours across two days) and have not managed to find any remotely concise way to trigger this. But then, I'm testing remotely on a PPC machine and not primarily a PPC developer so perhaps you will have a much better idea than I do about how to test here.</div><div><br></div><div>Anyways, replacing the SH operand to RLDICL here with the value '33' instead of '1' causes my high level tests to pass. But I can't generate useful test cases here so I don't feel comfortable committing that.</div><div><br></div><div>Anyways, I'll probably revert this until a test case can be added. Sorry for any disruption but hopefully the above analysis makes it easy to fix.</div></div></div><div><div class="gmail_quote"><div> <br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
+ return SDValue(CurDAG->getMachineNode(PPC::XORI, dl,<br>
+ MVT::i32, Shift, getI32Imm(1, dl)), 0);<br>
+ }<br>
}<br>
}<br>
<br>
@@ -2878,6 +2976,34 @@ SDValue PPCDAGToDAGISel::get32BitSExtCom<br>
getI32Imm(1, dl)), 0);<br>
return SDValue(CurDAG->getMachineNode(PPC::NEG, dl, MVT::i32, Xori), 0);<br>
}<br>
+ case ISD::SETGE: {<br>
+ // (sext (setcc %a, %b, setge)) -> (add (lshr (sub %a, %b), 63), -1)<br>
+ // (sext (setcc %a, 0, setge)) -> (ashr (~ %a), 31)<br>
+ if (IsRHSZero)<br>
+ return getCompoundZeroComparisonInGPR(LHS, dl, ZeroCompare::GESExt);<br>
+<br>
+ // Not a special case (i.e. RHS == 0). Handle (%a >= %b) as (%b <= %a)<br>
+ // by swapping inputs and falling through.<br>
+ std::swap(LHS, RHS);<br>
+ ConstantSDNode *RHSConst = dyn_cast<ConstantSDNode>(RHS);<br>
+ IsRHSZero = RHSConst && RHSConst->isNullValue();<br>
+ LLVM_FALLTHROUGH;<br>
+ }<br>
+ case ISD::SETLE: {<br>
+ // (sext (setcc %a, %b, setge)) -> (add (lshr (sub %b, %a), 63), -1)<br>
+ // (sext (setcc %a, 0, setle)) -> (add (lshr (- %a), 63), -1)<br>
+ if (IsRHSZero)<br>
+ return getCompoundZeroComparisonInGPR(LHS, dl, ZeroCompare::LESExt);<br>
+ SDValue SUBFNode =<br>
+ SDValue(CurDAG->getMachineNode(PPC::SUBF, dl, MVT::i32, MVT::Glue,<br>
+ LHS, RHS), 0);<br>
+ SDValue Srdi =<br>
+ SDValue(CurDAG->getMachineNode(PPC::RLDICL_32, dl, MVT::i32,<br>
+ SUBFNode, getI64Imm(1, dl),<br>
+ getI64Imm(63, dl)), 0);<br>
+ return SDValue(CurDAG->getMachineNode(PPC::ADDI, dl, MVT::i32, Srdi,<br>
+ getI32Imm(-1, dl)), 0);<br>
+ }<br>
}<br>
}<br>
<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesigesc.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesigesc.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesigesc.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesigesc.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesigesc.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i8 0, align 1<br>
+<br>
+define signext i32 @test_igesc(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_igesc:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %conv2 = zext i1 %cmp to i32<br>
+ ret i32 %conv2<br>
+}<br>
+<br>
+define signext i32 @test_igesc_sext(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_igesc_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ ret i32 %sub<br>
+}<br>
+<br>
+define void @test_igesc_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_igesc_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %conv3 = zext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_igesc_sext_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_igesc_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %conv3 = sext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesigesi.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesigesi.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesigesi.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesigesi.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesigesi.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i32 0, align 4<br>
+<br>
+define signext i32 @test_igesi(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_igesi:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %conv = zext i1 %cmp to i32<br>
+ ret i32 %conv<br>
+}<br>
+<br>
+define signext i32 @test_igesi_sext(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_igesi_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ ret i32 %sub<br>
+}<br>
+<br>
+define void @test_igesi_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_igesi_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %conv = zext i1 %cmp to i32<br>
+ store i32 %conv, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_igesi_sext_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_igesi_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ store i32 %sub, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesigess.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesigess.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesigess.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesigess.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesigess.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i16 0, align 2<br>
+<br>
+define signext i32 @test_igess(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_igess:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %conv2 = zext i1 %cmp to i32<br>
+ ret i32 %conv2<br>
+}<br>
+<br>
+define signext i32 @test_igess_sext(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_igess_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ ret i32 %sub<br>
+}<br>
+<br>
+define void @test_igess_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_igess_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %conv3 = zext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_igess_sext_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_igess_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %conv3 = sext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesilesc.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesilesc.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesilesc.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesilesc.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesilesc.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i8 0, align 1<br>
+<br>
+define signext i32 @test_ilesc(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_ilesc:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %conv2 = zext i1 %cmp to i32<br>
+ ret i32 %conv2<br>
+}<br>
+<br>
+define signext i32 @test_ilesc_sext(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_ilesc_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ ret i32 %sub<br>
+}<br>
+<br>
+define void @test_ilesc_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_ilesc_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %conv3 = zext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_ilesc_sext_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_ilesc_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %conv3 = sext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesilesi.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesilesi.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesilesi.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesilesi.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesilesi.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i32 0, align 4<br>
+<br>
+define signext i32 @test_ilesi(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_ilesi:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %conv = zext i1 %cmp to i32<br>
+ ret i32 %conv<br>
+}<br>
+<br>
+define signext i32 @test_ilesi_sext(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_ilesi_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ ret i32 %sub<br>
+}<br>
+<br>
+define void @test_ilesi_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_ilesi_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %conv = zext i1 %cmp to i32<br>
+ store i32 %conv, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_ilesi_sext_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_ilesi_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ store i32 %sub, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesiless.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesiless.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesiless.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesiless.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesiless.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i16 0, align 2<br>
+<br>
+define signext i32 @test_iless(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_iless:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %conv2 = zext i1 %cmp to i32<br>
+ ret i32 %conv2<br>
+}<br>
+<br>
+define signext i32 @test_iless_sext(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_iless_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ ret i32 %sub<br>
+}<br>
+<br>
+define void @test_iless_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_iless_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %conv3 = zext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_iless_sext_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_iless_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %conv3 = sext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesllgesc.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllgesc.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllgesc.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesllgesc.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesllgesc.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i8 0, align 1<br>
+<br>
+define i64 @test_llgesc(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_llgesc:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %conv3 = zext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define i64 @test_llgesc_sext(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_llgesc_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %conv3 = sext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define void @test_llgesc_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_llgesc_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %conv3 = zext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_llgesc_sext_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_llgesc_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i8 %a, %b<br>
+ %conv3 = sext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesllgesi.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllgesi.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllgesi.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesllgesi.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesllgesi.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i32 0, align 4<br>
+<br>
+define i64 @test_llgesi(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_llgesi:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %conv1 = zext i1 %cmp to i64<br>
+ ret i64 %conv1<br>
+}<br>
+<br>
+define i64 @test_llgesi_sext(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_llgesi_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %conv1 = sext i1 %cmp to i64<br>
+ ret i64 %conv1<br>
+}<br>
+<br>
+define void @test_llgesi_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_llgesi_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %conv = zext i1 %cmp to i32<br>
+ store i32 %conv, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_llgesi_sext_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_llgesi_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i32 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ store i32 %sub, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesllgess.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllgess.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllgess.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesllgess.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesllgess.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,68 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+@glob = common local_unnamed_addr global i16 0, align 2<br>
+<br>
+define i64 @test_llgess(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llgess:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %conv3 = zext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define i64 @test_llgess_sext(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llgess_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %conv3 = sext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define void @test_llgess_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llgess_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %conv3 = zext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_llgess_sext_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llgess_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r4, r3<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sge i16 %a, %b<br>
+ %conv3 = sext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testCompareslllesc.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testCompareslllesc.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testCompareslllesc.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testCompareslllesc.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testCompareslllesc.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,69 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+<br>
+@glob = common local_unnamed_addr global i8 0, align 1<br>
+<br>
+define i64 @test_lllesc(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_lllesc:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %conv3 = zext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define i64 @test_lllesc_sext(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_lllesc_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %conv3 = sext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define void @test_lllesc_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_lllesc_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %conv3 = zext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_lllesc_sext_store(i8 signext %a, i8 signext %b) {<br>
+; CHECK-LABEL: test_lllesc_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stb r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i8 %a, %b<br>
+ %conv3 = sext i1 %cmp to i8<br>
+ store i8 %conv3, i8* @glob, align 1<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testCompareslllesi.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testCompareslllesi.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testCompareslllesi.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testCompareslllesi.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testCompareslllesi.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,69 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+<br>
+@glob = common local_unnamed_addr global i32 0, align 4<br>
+<br>
+define i64 @test_lllesi(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_lllesi:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %conv1 = zext i1 %cmp to i64<br>
+ ret i64 %conv1<br>
+}<br>
+<br>
+define i64 @test_lllesi_sext(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_lllesi_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %conv1 = sext i1 %cmp to i64<br>
+ ret i64 %conv1<br>
+}<br>
+<br>
+define void @test_lllesi_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_lllesi_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %conv = zext i1 %cmp to i32<br>
+ store i32 %conv, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_lllesi_sext_store(i32 signext %a, i32 signext %b) {<br>
+; CHECK-LABEL: test_lllesi_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: stw r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i32 %a, %b<br>
+ %sub = sext i1 %cmp to i32<br>
+ store i32 %sub, i32* @glob, align 4<br>
+ ret void<br>
+}<br>
<br>
Added: llvm/trunk/test/CodeGen/PowerPC/testComparesllless.ll<br>
URL: <a href="http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllless.ll?rev=310346&view=auto" rel="noreferrer" target="_blank">http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/PowerPC/testComparesllless.ll?rev=310346&view=auto</a><br>
==============================================================================<br>
--- llvm/trunk/test/CodeGen/PowerPC/testComparesllless.ll (added)<br>
+++ llvm/trunk/test/CodeGen/PowerPC/testComparesllless.ll Tue Aug 8 04:20:44 2017<br>
@@ -0,0 +1,69 @@<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu -O2 \<br>
+; RUN: -ppc-asm-full-reg-names -mcpu=pwr8 < %s | FileCheck %s \<br>
+; RUN: --implicit-check-not cmpw --implicit-check-not cmpd --implicit-check-not cmpl<br>
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py<br>
+<br>
+@glob = common local_unnamed_addr global i16 0, align 2<br>
+<br>
+define i64 @test_llless(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llless:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %conv3 = zext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define i64 @test_llless_sext(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llless_sext:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %conv3 = sext i1 %cmp to i64<br>
+ ret i64 %conv3<br>
+}<br>
+<br>
+define void @test_llless_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llless_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: xori r3, r3, 1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %conv3 = zext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
+<br>
+define void @test_llless_sext_store(i16 signext %a, i16 signext %b) {<br>
+; CHECK-LABEL: test_llless_sext_store:<br>
+; CHECK: # BB#0: # %entry<br>
+; CHECK-NEXT: addis r5, r2, .LC0@toc@ha<br>
+; CHECK-NEXT: subf r3, r3, r4<br>
+; CHECK-NEXT: ld r12, .LC0@toc@l(r5)<br>
+; CHECK-NEXT: rldicl r3, r3, 1, 63<br>
+; CHECK-NEXT: addi r3, r3, -1<br>
+; CHECK-NEXT: sth r3, 0(r12)<br>
+; CHECK-NEXT: blr<br>
+entry:<br>
+ %cmp = icmp sle i16 %a, %b<br>
+ %conv3 = sext i1 %cmp to i16<br>
+ store i16 %conv3, i16* @glob, align 2<br>
+ ret void<br>
+}<br>
<br>
<br>
_______________________________________________<br>
llvm-commits mailing list<br>
<a href="mailto:llvm-commits@lists.llvm.org" target="_blank">llvm-commits@lists.llvm.org</a><br>
<a href="http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits" rel="noreferrer" target="_blank">http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits</a><br>
</blockquote></div></div></blockquote></div>
</div></div></blockquote></div><br></div>
</blockquote></div></div></blockquote></div></div></div></div></blockquote></div><br></div>
</blockquote></div></div>