Re: [meta-arm] [PATCH] arm-toolchain: Fix potential runtime crash

["Denys Dmytriyenko" <denis@denix.org>]

Jon,

I'm fine with backporting a patch from upstream for gcc 9. It should probably 
be merged to Dunfell as well. And master can drop gcc 9 recipes for building 
from sources to match oe-core master, which only has gcc 10.

-- 
Denys


On Fri, Feb 12, 2021 at 10:22:16AM -0500, Jon Mason wrote:
> Unless you have a problem with the patch below, I'm inclined to pull
> it in ASAP.  However, given that the Arm GCC 9.x has been EoL'ed in
> favor of the 10.x code, should we just drop support for it?
> 
> Thanks,
> Jon
> 
> ----- Forwarded message from Jon Mason <jon.mason@arm.com> -----
> 
> Date: Thu, 11 Feb 2021 11:36:47 -0500
> From: Jon Mason <jon.mason@arm.com>
> To: meta-arm@lists.yoctoproject.org
> Subject: [meta-arm] [PATCH] arm-toolchain: Fix potential runtime crash
> 
> GCCv9 tree vectorization code is faulty and can cause random crashes at
> runtime (when using -O3).  Add the backported patch to address this
> issue.
> 
> Change-Id: If7bb0ba0720bab42e7d34f3679d988934f657392
> Signed-off-by: Jon Mason <jon.mason@arm.com>
> ---
>  .../recipes-devtools/gcc/gcc-arm-9.2.inc      |   1 +
>  ...-PR-tree-optimization-97236-fix-bad-.patch | 119 ++++++++++++++++++
>  2 files changed, 120 insertions(+)
>  create mode 100644 meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2/0001-Backport-fix-for-PR-tree-optimization-97236-fix-bad-.patch
> 
> diff --git a/meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2.inc b/meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2.inc
> index 08ad796..6378ecf 100644
> --- a/meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2.inc
> +++ b/meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2.inc
> @@ -68,6 +68,7 @@ SRC_URI = "\
>             file://0035-Fix-for-testsuite-failure.patch \
>             file://0036-Re-introduce-spe-commandline-options.patch \
>             file://0037-Fix-up-libsanitizer-build-with-master-glibc.patch \
> +           file://0001-Backport-fix-for-PR-tree-optimization-97236-fix-bad-.patch \
>  "
>  SRC_URI[md5sum] = "9c570fc4286825b4e6f67b3d34aade23"
>  
> diff --git a/meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2/0001-Backport-fix-for-PR-tree-optimization-97236-fix-bad-.patch b/meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2/0001-Backport-fix-for-PR-tree-optimization-97236-fix-bad-.patch
> new file mode 100644
> index 0000000..dc1039d
> --- /dev/null
> +++ b/meta-arm-toolchain/recipes-devtools/gcc/gcc-arm-9.2/0001-Backport-fix-for-PR-tree-optimization-97236-fix-bad-.patch
> @@ -0,0 +1,119 @@
> +Upstream-Status: Backport [https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=97b668f9a8c6ec565c278a60e7d1492a6932e409]
> +Signed-off-by: Jon Mason <jon.mason@arm.com>
> +
> +From 97b668f9a8c6ec565c278a60e7d1492a6932e409 Mon Sep 17 00:00:00 2001
> +From: Matthias Klose <doko@ubuntu.com>
> +Date: Tue, 6 Oct 2020 13:41:37 +0200
> +Subject: [PATCH] Backport fix for PR/tree-optimization/97236 - fix bad use of
> + VMAT_CONTIGUOUS
> +
> +This avoids using VMAT_CONTIGUOUS with single-element interleaving
> +when using V1mode vectors.  Instead keep VMAT_ELEMENTWISE but
> +continue to avoid load-lanes and gathers.
> +
> +2020-10-01  Richard Biener  <rguenther@suse.de>
> +
> +	PR tree-optimization/97236
> +	* tree-vect-stmts.c (get_group_load_store_type): Keep
> +	VMAT_ELEMENTWISE for single-element vectors.
> +
> +	* gcc.dg/vect/pr97236.c: New testcase.
> +
> +(cherry picked from commit 1ab88985631dd2c5a5e3b5c0dce47cf8b6ed2f82)
> +---
> + gcc/testsuite/gcc.dg/vect/pr97236.c | 43 +++++++++++++++++++++++++++++
> + gcc/tree-vect-stmts.c               | 20 ++++++--------
> + 2 files changed, 52 insertions(+), 11 deletions(-)
> + create mode 100644 gcc/testsuite/gcc.dg/vect/pr97236.c
> +
> +diff --git a/gcc/testsuite/gcc.dg/vect/pr97236.c b/gcc/testsuite/gcc.dg/vect/pr97236.c
> +new file mode 100644
> +index 000000000000..9d3dc20d953d
> +--- /dev/null
> ++++ b/gcc/testsuite/gcc.dg/vect/pr97236.c
> +@@ -0,0 +1,43 @@
> ++typedef unsigned char __uint8_t;
> ++typedef __uint8_t uint8_t;
> ++typedef struct plane_t {
> ++  uint8_t *p_pixels;
> ++  int i_lines;
> ++  int i_pitch;
> ++} plane_t;
> ++
> ++typedef struct {
> ++  plane_t p[5];
> ++} picture_t;
> ++
> ++#define N 4
> ++
> ++void __attribute__((noipa))
> ++picture_Clone(picture_t *picture, picture_t *res)
> ++{
> ++  for (int i = 0; i < N; i++) {
> ++    res->p[i].p_pixels = picture->p[i].p_pixels;
> ++    res->p[i].i_lines = picture->p[i].i_lines;
> ++    res->p[i].i_pitch = picture->p[i].i_pitch;
> ++  }
> ++}
> ++
> ++int
> ++main()
> ++{
> ++  picture_t aaa, bbb;
> ++  uint8_t pixels[10] = {1, 1, 1, 1, 1, 1, 1, 1};
> ++
> ++  for (unsigned i = 0; i < N; i++)
> ++    aaa.p[i].p_pixels = pixels;
> ++
> ++  picture_Clone (&aaa, &bbb);
> ++
> ++  uint8_t c = 0;
> ++  for (unsigned i = 0; i < N; i++)
> ++    c += bbb.p[i].p_pixels[0];
> ++
> ++  if (c != N)
> ++    __builtin_abort ();
> ++  return 0;
> ++}
> +diff --git a/gcc/tree-vect-stmts.c b/gcc/tree-vect-stmts.c
> +index 507f81b0a0e8..ffbba3441de2 100644
> +--- a/gcc/tree-vect-stmts.c
> ++++ b/gcc/tree-vect-stmts.c
> +@@ -2355,25 +2355,23 @@ get_group_load_store_type (stmt_vec_info stmt_info, tree vectype, bool slp,
> + 	  /* First cope with the degenerate case of a single-element
> + 	     vector.  */
> + 	  if (known_eq (TYPE_VECTOR_SUBPARTS (vectype), 1U))
> +-	    *memory_access_type = VMAT_CONTIGUOUS;
> ++	    ;
> + 
> + 	  /* Otherwise try using LOAD/STORE_LANES.  */
> +-	  if (*memory_access_type == VMAT_ELEMENTWISE
> +-	      && (vls_type == VLS_LOAD
> +-		  ? vect_load_lanes_supported (vectype, group_size, masked_p)
> +-		  : vect_store_lanes_supported (vectype, group_size,
> +-						masked_p)))
> ++	  else if (vls_type == VLS_LOAD
> ++		   ? vect_load_lanes_supported (vectype, group_size, masked_p)
> ++		   : vect_store_lanes_supported (vectype, group_size,
> ++						 masked_p))
> + 	    {
> + 	      *memory_access_type = VMAT_LOAD_STORE_LANES;
> + 	      overrun_p = would_overrun_p;
> + 	    }
> + 
> + 	  /* If that fails, try using permuting loads.  */
> +-	  if (*memory_access_type == VMAT_ELEMENTWISE
> +-	      && (vls_type == VLS_LOAD
> +-		  ? vect_grouped_load_supported (vectype, single_element_p,
> +-						 group_size)
> +-		  : vect_grouped_store_supported (vectype, group_size)))
> ++	  else if (vls_type == VLS_LOAD
> ++		   ? vect_grouped_load_supported (vectype, single_element_p,
> ++						  group_size)
> ++		   : vect_grouped_store_supported (vectype, group_size))
> + 	    {
> + 	      *memory_access_type = VMAT_CONTIGUOUS_PERMUTE;
> + 	      overrun_p = would_overrun_p;
> +-- 
> +2.20.1
> +
> -- 
> 2.17.1
> 
> 
> 
> 
> 
> 
> 
> ----- End forwarded message -----

> 
> 
> 


-- 
Regards,
Denys Dmytriyenko <denis@denix.org>
PGP: 0x420902729A92C964 - https://denix.org/0x420902729A92C964
Fingerprint: 25FC E4A5 8A72 2F69 1186  6D76 4209 0272 9A92 C964