[LoongArch] Fix GOT usage for non-dso_local function calls in large code model

This commit fixes an issue in the large code model where non-dso_local
function calls did not use the GOT as expected in PIC mode. Instead,
direct PC-relative access was incorrectly applied, leading to linker
errors when building shared libraries.

For `ExternalSymbol`, it is not possible to determine whether it is
dso_local during pseudo-instruction expansion. We use target flags to
differentiate whether GOT should be used.

Cherry-picked from #117099, used for fix linker errors when bulding
shared libraries with large code model.

Author: wangleiat

llvm/lib/Target/LoongArch/LoongArchExpandPseudoInsts.cpp

@@ -721,7 +721,7 @@ bool LoongArchExpandPseudo::expandFunctionCALL(
         IsTailCall ? LoongArch::PseudoJIRL_TAIL : LoongArch::PseudoJIRL_CALL;
     Register AddrReg = IsTailCall ? LoongArch::R19 : LoongArch::R1;
 
-    bool UseGOT = Func.isGlobal() && !Func.getGlobal()->isDSOLocal();
+    bool UseGOT = Func.getTargetFlags() == LoongArchII::MO_CALL_PLT;
     unsigned MO = UseGOT ? LoongArchII::MO_GOT_PC_HI : LoongArchII::MO_PCREL_LO;
     unsigned LAOpcode = UseGOT ? LoongArch::LDX_D : LoongArch::ADD_D;
     expandLargeAddressLoad(MBB, MBBI, NextMBBI, LAOpcode, MO, Func, AddrReg,

llvm/test/CodeGen/LoongArch/code-models.ll

@@ -82,11 +82,11 @@ define void @call_external_sym(ptr %dst) {
 ; LARGE-NEXT:    .cfi_offset 1, -8
 ; LARGE-NEXT:    ori $a2, $zero, 1000
 ; LARGE-NEXT:    move $a1, $zero
-; LARGE-NEXT:    pcalau12i $ra, %pc_hi20(memset)
-; LARGE-NEXT:    addi.d $t8, $zero, %pc_lo12(memset)
-; LARGE-NEXT:    lu32i.d $t8, %pc64_lo20(memset)
-; LARGE-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(memset)
-; LARGE-NEXT:    add.d $ra, $t8, $ra
+; LARGE-NEXT:    pcalau12i $ra, %got_pc_hi20(memset)
+; LARGE-NEXT:    addi.d $t8, $zero, %got_pc_lo12(memset)
+; LARGE-NEXT:    lu32i.d $t8, %got64_pc_lo20(memset)
+; LARGE-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(memset)
+; LARGE-NEXT:    ldx.d $ra, $t8, $ra
 ; LARGE-NEXT:    jirl $ra, $ra, 0
 ; LARGE-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
 ; LARGE-NEXT:    addi.d $sp, $sp, 16

llvm/test/CodeGen/LoongArch/machinelicm-address-pseudos.ll

@@ -282,11 +282,11 @@ define void @test_la_tls_ld(i32 signext %n) {
 ; LA64LARGE-NEXT:  .LBB3_1: # %loop
 ; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
 ; LA64LARGE-NEXT:    move $a0, $s0
-; LA64LARGE-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LA64LARGE-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LA64LARGE-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LA64LARGE-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LA64LARGE-NEXT:    add.d $ra, $t8, $ra
+; LA64LARGE-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LA64LARGE-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LA64LARGE-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LA64LARGE-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LA64LARGE-NEXT:    ldx.d $ra, $t8, $ra
 ; LA64LARGE-NEXT:    jirl $ra, $ra, 0
 ; LA64LARGE-NEXT:    ld.w $zero, $a0, 0
 ; LA64LARGE-NEXT:    addi.w $s1, $s1, 1
@@ -448,11 +448,11 @@ define void @test_la_tls_gd(i32 signext %n) nounwind {
 ; LA64LARGE-NEXT:  .LBB5_1: # %loop
 ; LA64LARGE-NEXT:    # =>This Inner Loop Header: Depth=1
 ; LA64LARGE-NEXT:    move $a0, $s0
-; LA64LARGE-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LA64LARGE-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LA64LARGE-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LA64LARGE-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LA64LARGE-NEXT:    add.d $ra, $t8, $ra
+; LA64LARGE-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LA64LARGE-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LA64LARGE-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LA64LARGE-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LA64LARGE-NEXT:    ldx.d $ra, $t8, $ra
 ; LA64LARGE-NEXT:    jirl $ra, $ra, 0
 ; LA64LARGE-NEXT:    ld.w $zero, $a0, 0
 ; LA64LARGE-NEXT:    addi.w $s1, $s1, 1

llvm/test/CodeGen/LoongArch/psabi-restricted-scheduling.ll

@@ -105,23 +105,23 @@ define void @foo() nounwind {
 ; LARGE_NO_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(gd)
 ; LARGE_NO_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(gd)
 ; LARGE_NO_SCH-NEXT:    add.d $a0, $t8, $a0
-; LARGE_NO_SCH-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    add.d $ra, $t8, $ra
+; LARGE_NO_SCH-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    ldx.d $ra, $t8, $ra
 ; LARGE_NO_SCH-NEXT:    jirl $ra, $ra, 0
 ; LARGE_NO_SCH-NEXT:    ld.d $zero, $a0, 0
 ; LARGE_NO_SCH-NEXT:    pcalau12i $a0, %ld_pc_hi20(ld)
 ; LARGE_NO_SCH-NEXT:    addi.d $t8, $zero, %got_pc_lo12(ld)
 ; LARGE_NO_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(ld)
 ; LARGE_NO_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(ld)
 ; LARGE_NO_SCH-NEXT:    add.d $a0, $t8, $a0
-; LARGE_NO_SCH-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LARGE_NO_SCH-NEXT:    add.d $ra, $t8, $ra
+; LARGE_NO_SCH-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LARGE_NO_SCH-NEXT:    ldx.d $ra, $t8, $ra
 ; LARGE_NO_SCH-NEXT:    jirl $ra, $ra, 0
 ; LARGE_NO_SCH-NEXT:    pcalau12i $a1, %ie_pc_hi20(ie)
 ; LARGE_NO_SCH-NEXT:    addi.d $t8, $zero, %ie_pc_lo12(ie)
@@ -162,23 +162,23 @@ define void @foo() nounwind {
 ; LARGE_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(gd)
 ; LARGE_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(gd)
 ; LARGE_SCH-NEXT:    add.d $a0, $t8, $a0
-; LARGE_SCH-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LARGE_SCH-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LARGE_SCH-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LARGE_SCH-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LARGE_SCH-NEXT:    add.d $ra, $t8, $ra
+; LARGE_SCH-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LARGE_SCH-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LARGE_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LARGE_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LARGE_SCH-NEXT:    ldx.d $ra, $t8, $ra
 ; LARGE_SCH-NEXT:    jirl $ra, $ra, 0
 ; LARGE_SCH-NEXT:    ld.d $zero, $a0, 0
 ; LARGE_SCH-NEXT:    pcalau12i $a0, %ld_pc_hi20(ld)
 ; LARGE_SCH-NEXT:    addi.d $t8, $zero, %got_pc_lo12(ld)
 ; LARGE_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(ld)
 ; LARGE_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(ld)
 ; LARGE_SCH-NEXT:    add.d $a0, $t8, $a0
-; LARGE_SCH-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LARGE_SCH-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LARGE_SCH-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LARGE_SCH-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LARGE_SCH-NEXT:    add.d $ra, $t8, $ra
+; LARGE_SCH-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LARGE_SCH-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LARGE_SCH-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LARGE_SCH-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LARGE_SCH-NEXT:    ldx.d $ra, $t8, $ra
 ; LARGE_SCH-NEXT:    jirl $ra, $ra, 0
 ; LARGE_SCH-NEXT:    pcalau12i $a1, %ie_pc_hi20(ie)
 ; LARGE_SCH-NEXT:    addi.d $t8, $zero, %ie_pc_lo12(ie)

llvm/test/CodeGen/LoongArch/tls-models.ll

@@ -55,11 +55,11 @@ define ptr @f1() nounwind {
 ; LA64LARGEPIC-NEXT:    lu32i.d $t8, %got64_pc_lo20(unspecified)
 ; LA64LARGEPIC-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(unspecified)
 ; LA64LARGEPIC-NEXT:    add.d $a0, $t8, $a0
-; LA64LARGEPIC-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    add.d $ra, $t8, $ra
+; LA64LARGEPIC-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    ldx.d $ra, $t8, $ra
 ; LA64LARGEPIC-NEXT:    jirl $ra, $ra, 0
 ; LA64LARGEPIC-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
 ; LA64LARGEPIC-NEXT:    addi.d $sp, $sp, 16
@@ -169,11 +169,11 @@ define ptr @f2() nounwind {
 ; LA64LARGEPIC-NEXT:    lu32i.d $t8, %got64_pc_lo20(ld)
 ; LA64LARGEPIC-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(ld)
 ; LA64LARGEPIC-NEXT:    add.d $a0, $t8, $a0
-; LA64LARGEPIC-NEXT:    pcalau12i $ra, %pc_hi20(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    addi.d $t8, $zero, %pc_lo12(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    lu32i.d $t8, %pc64_lo20(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    lu52i.d $t8, $t8, %pc64_hi12(__tls_get_addr)
-; LA64LARGEPIC-NEXT:    add.d $ra, $t8, $ra
+; LA64LARGEPIC-NEXT:    pcalau12i $ra, %got_pc_hi20(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    addi.d $t8, $zero, %got_pc_lo12(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    lu32i.d $t8, %got64_pc_lo20(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    lu52i.d $t8, $t8, %got64_pc_hi12(__tls_get_addr)
+; LA64LARGEPIC-NEXT:    ldx.d $ra, $t8, $ra
 ; LA64LARGEPIC-NEXT:    jirl $ra, $ra, 0
 ; LA64LARGEPIC-NEXT:    ld.d $ra, $sp, 8 # 8-byte Folded Reload
 ; LA64LARGEPIC-NEXT:    addi.d $sp, $sp, 16