mirror of
https://github.com/llvm/llvm-project.git
synced 2025-04-27 05:56:08 +00:00
149f5b573c
This is an alternate fix (see D87835) for a bug where a NaN constant gets wrongly transformed into Infinity via truncation. In this patch, we uniformly convert any SNaN to QNaN while raising 'invalid op'. But we don't have a way to directly specify a 32-bit SNaN value in LLVM IR, so those are always encoded/decoded by calling convert from/to 64-bit hex. See D88664 for a clang fix needed to allow this change. Differential Revision: https://reviews.llvm.org/D88238
60 lines
3.5 KiB
C
60 lines
3.5 KiB
C
// RUN: %clang -target mipsel-unknown-linux -mnan=2008 -march=mips2 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NANLEGACY %s
|
|
// RUN: FileCheck -check-prefix=CHECK-MIPS2 %s < %t
|
|
//
|
|
// RUN: %clang -target mips64el-unknown-linux -mnan=2008 -march=mips3 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NANLEGACY %s
|
|
// RUN: FileCheck -check-prefix=CHECK-MIPS3 %s < %t
|
|
//
|
|
// RUN: %clang -target mips64el-unknown-linux -mnan=2008 -march=mips4 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NANLEGACY %s
|
|
// RUN: FileCheck -check-prefix=CHECK-MIPS4 %s < %t
|
|
//
|
|
// RUN: %clang -target mipsel-unknown-linux -mnan=2008 -march=mips32 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NANLEGACY %s
|
|
// RUN: FileCheck -check-prefix=CHECK-MIPS32 %s < %t
|
|
//
|
|
// RUN: %clang -target mipsel-unknown-linux -mnan=2008 -march=mips32r2 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NAN2008 %s
|
|
// RUN: FileCheck -allow-empty -check-prefix=NO-WARNINGS %s < %t
|
|
//
|
|
// RUN: %clang -target mipsel-unknown-linux -mnan=2008 -march=mips32r3 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NAN2008 %s
|
|
// RUN: FileCheck -allow-empty -check-prefix=NO-WARNINGS %s < %t
|
|
//
|
|
// RUN: %clang -target mipsel-unknown-linux -mnan=legacy -march=mips32r6 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NAN2008 %s
|
|
// RUN: FileCheck -check-prefix=CHECK-MIPS32R6 %s < %t
|
|
//
|
|
// RUN: %clang -target mips64el-unknown-linux -mnan=2008 -march=mips64 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NANLEGACY %s
|
|
// RUN: FileCheck -check-prefix=CHECK-MIPS64 %s < %t
|
|
//
|
|
// RUN: %clang -target mips64el-unknown-linux -mnan=2008 -march=mips64r2 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NAN2008 %s
|
|
// RUN: FileCheck -allow-empty -check-prefix=NO-WARNINGS %s < %t
|
|
//
|
|
// RUN: %clang -target mips64el-unknown-linux -mnan=legacy -march=mips64r6 -emit-llvm -S %s -o - 2>%t | FileCheck -check-prefix=CHECK-NAN2008 %s
|
|
// RUN: FileCheck -check-prefix=CHECK-MIPS64R6 %s < %t
|
|
|
|
// NO-WARNINGS-NOT: warning: ignoring '-mnan=legacy' option
|
|
// NO-WARNINGS-NOT: warning: ignoring '-mnan=2008' option
|
|
|
|
// CHECK-MIPS2: warning: ignoring '-mnan=2008' option because the 'mips2' architecture does not support it
|
|
// CHECK-MIPS3: warning: ignoring '-mnan=2008' option because the 'mips3' architecture does not support it
|
|
// CHECK-MIPS4: warning: ignoring '-mnan=2008' option because the 'mips4' architecture does not support it
|
|
// CHECK-MIPS32: warning: ignoring '-mnan=2008' option because the 'mips32' architecture does not support it
|
|
// CHECK-MIPS32R6: warning: ignoring '-mnan=legacy' option because the 'mips32r6' architecture does not support it
|
|
// CHECK-MIPS64: warning: ignoring '-mnan=2008' option because the 'mips64' architecture does not support it
|
|
// CHECK-MIPS64R6: warning: ignoring '-mnan=legacy' option because the 'mips64r6' architecture does not support it
|
|
|
|
// This call creates a QNAN double with an empty payload.
|
|
// The quiet bit is inverted in legacy mode: it is clear to indicate QNAN,
|
|
// so the next highest bit is set to maintain NAN (not infinity).
|
|
// In regular (2008) mode, the quiet bit is set to indicate QNAN.
|
|
|
|
// CHECK-NANLEGACY: double 0x7FF4000000000000
|
|
// CHECK-NAN2008: double 0x7FF8000000000000
|
|
|
|
double d = __builtin_nan("");
|
|
|
|
// This call creates a QNAN double with an empty payload and then truncates.
|
|
// llvm::APFloat does not know about the inverted quiet bit, so it sets the
|
|
// quiet bit on conversion independently of the setting in clang.
|
|
|
|
// CHECK-NANLEGACY: float 0x7FFC000000000000
|
|
// CHECK-NAN2008: float 0x7FF8000000000000
|
|
|
|
float f = __builtin_nan("");
|