Delete deprecated fp32 precision warnings (pytorch#166956) (pytorch#166998)

The deprecation warning led to warning spamming in PyTorch APIs, like
torch.compile. This is not how a deprecation warning should go: if we
add a deprecation warning, we'd better update our built-in APIs to
prevent warning spam.
Pull Request resolved: pytorch#166956
Approved by: https://github.com/albanD

(cherry picked from commit 527b110)

Co-authored-by: Richard Zou <[email protected]>

Author: Lucaskabela

aten/src/ATen/Context.cpp

@@ -24,7 +24,6 @@ C10_DIAGNOSTIC_POP()
 namespace at {
 
 namespace {
-
 /*
   These const variables defined the fp32 precisions for different backend
   We have "generic", "cuda", "mkldnn" backend now and we can choose fp32
@@ -76,14 +75,6 @@ void check_fp32_prec_backend_and_op(
     return valid;
   }
 
-  C10_ALWAYS_INLINE void warn_deprecated_fp32_precision_api(){
-    TORCH_WARN_ONCE(
-      "Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' "
-      "or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, "
-      "torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see "
-      "https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices"
-    );
-  }
 } // namespace
 
 Context::Context() = default;
@@ -193,15 +184,13 @@ bool Context::allowTF32CuDNN(const std::string& op) const {
   } else {
     return float32Precision("cuda", op) == "tf32";
   }
-  warn_deprecated_fp32_precision_api();
   return allow_tf32_cudnn;
 }
 
 void Context::setAllowTF32CuDNN(bool b) {
   setFloat32Precision("cuda", "rnn", b ? "tf32" : "none");
   setFloat32Precision("cuda", "conv", b ? "tf32" : "none");
   allow_tf32_cudnn = b;
-  warn_deprecated_fp32_precision_api();
 }
 
 void Context::setSDPPriorityOrder(const std::vector<int64_t>& order) {
@@ -357,7 +346,6 @@ bool Context::allowTF32CuBLAS() const {
       "Current status indicate that you have used mix of the legacy and new APIs to set the TF32 status for cublas matmul. ",
       "We suggest only using the new API to set the TF32 flag. See also: ",
       "https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices");
-  warn_deprecated_fp32_precision_api();
   return allow_tf32_new;
 }
 
@@ -389,7 +377,6 @@ Float32MatmulPrecision Context::float32MatmulPrecision() const {
       "Current status indicate that you have used mix of the legacy and new APIs to set the matmul precision. ",
       "We suggest only using the new API for matmul precision. See also: ",
       "https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices");
-  warn_deprecated_fp32_precision_api();
   return float32_matmul_precision;
 }
 
@@ -406,7 +393,6 @@ std::string Context::float32Precision(const std::string& backend, const std::str
 
 void Context::setFloat32MatmulPrecision(const std::string &s) {
   auto match = [this](const std::string & s_) {
-    warn_deprecated_fp32_precision_api();
     // TODO: consider if CuDNN field needs to also be set for potential future CuDNN ops like multi-headed attention
     if (s_ == "highest") {
       float32_matmul_precision = at::Float32MatmulPrecision::HIGHEST;