[arch-commits] Commit in python-pytorch/trunk (PKGBUILD fix_c10.patch)

[Konstantin Gizdov]

    Date: Wednesday, August 18, 2021 @ 02:24:26
  Author: kgizdov
Revision: 1002856

upgpkg: python-pytorch 1.9.0-5: cuda 11.4.1 rebuild

Added:
  python-pytorch/trunk/fix_c10.patch
Modified:
  python-pytorch/trunk/PKGBUILD

---------------+
 PKGBUILD      |   10 +++++++---
 fix_c10.patch |   12 ++++++++++++
 2 files changed, 19 insertions(+), 3 deletions(-)

Modified: PKGBUILD
===================================================================
--- PKGBUILD	2021-08-18 01:30:41 UTC (rev 1002855)
+++ PKGBUILD	2021-08-18 02:24:26 UTC (rev 1002856)
@@ -6,7 +6,7 @@
 _pkgname="pytorch"
 pkgver=1.9.0
 _pkgver=1.9.0
-pkgrel=4
+pkgrel=5
 pkgdesc="Tensors and Dynamic neural networks in Python with strong GPU acceleration"
 arch=('x86_64')
 url="https://pytorch.org"
@@ -58,6 +58,7 @@
         benchmark-gcc11.patch
         xnnpack-gcc11.patch
         https://github.com/pytorch/pytorch/commit/c74c0c571880df886474be297c556562e95c00e0.patch
+        fix_c10.patch
         disable_non_x86_64.patch)
 sha256sums=('SKIP'
             'SKIP'
@@ -99,7 +100,8 @@
             '689c76e89bcf403df1b4cf7ca784381967b6a6527ed6eb6d0ad6681cf789b738'
             '278fecdb45df065343f51688cc7a1665153b5189f3341a741d546b0b518eac40'
             '64833e96e47a22f88336381f25fcd73127208dc79e2074398295d88c4596c06a'
-            '3d5b9d3bbba3238d8f165e582039ec07798bccc1d1f44bd91e8b1892236cb70f'
+            'b106af479ef78803b39868fa1458c22c6e7fdbd13a15ebca20893e1c81c75463'
+            'ba801238afcfc58a35410e54d4ca6a638c447865c0c6b38ed16917fd6d507954'
             'd3ef8491718ed7e814fe63e81df2f49862fffbea891d2babbcb464796a1bd680')
 
 prepare() {
@@ -160,6 +162,8 @@
 
   # cuda 11.4 fix
   patch -Np1 <../c74c0c571880df886474be297c556562e95c00e0.patch
+  # cuda 11.4.1 fix
+  patch -Np1 -i "${srcdir}/fix_c10.patch"
 
   # remove local nccl
   rm -rf third_party/nccl/nccl
@@ -196,7 +200,7 @@
   # export BUILD_SPLIT_CUDA=ON  # modern preferred build, but splits libs and symbols, ABI break
   # export USE_FAST_NVCC=ON  # parallel build with nvcc, spawns too many processes
   export USE_CUPTI_SO=ON  # make sure cupti.so is used as shared lib
-  export CUDAHOSTCXX=/usr/bin/g++-10
+  export CUDAHOSTCXX=/usr/bin/g++
   export CUDA_HOST_COMPILER="${CUDAHOSTCXX}"
   export CUDA_HOME=/opt/cuda
   export CUDNN_LIB_DIR=/usr/lib

Added: fix_c10.patch
===================================================================
--- fix_c10.patch	                        (rev 0)
+++ fix_c10.patch	2021-08-18 02:24:26 UTC (rev 1002856)
@@ -0,0 +1,12 @@
+diff --color -aur pytorch-1.9.0-old/c10/core/TensorImpl.h pytorch-1.9.0-new/c10/core/TensorImpl.h
+--- pytorch-1.9.0-old/c10/core/TensorImpl.h	2021-08-17 19:33:40.324974399 +0300
++++ pytorch-1.9.0-new/c10/core/TensorImpl.h	2021-08-18 01:25:00.005901707 +0300
+@@ -2177,7 +2177,7 @@
+ //    DispatchKeySet
+ //
+ static_assert(
+-    sizeof(void*) != sizeof(int64_t) || // if 64-bit...
++    sizeof(void*) <= sizeof(int64_t) || // if 64-bit...
+         sizeof(TensorImpl) == sizeof(int64_t) * 23,
+     "You changed the size of TensorImpl on 64-bit arch."
+     "See Note [TensorImpl size constraints] on how to proceed.");