RE: Crashes on ppc/s390 after 8231441: AArch64: Initial SVE backend support

[Ningsheng Jian]

Hi Goetz,

I am sorry about that and thanks for helping to identify the issue.

As I cannot reproduce this on x86 and AArch64, a quick guess is that I may have 
missed the code like:

diff --git a/src/hotspot/share/opto/type.cpp b/src/hotspot/share/opto/type.cpp
index 2f4047dfa8e..f7d2f5b2320 100644
--- a/src/hotspot/share/opto/type.cpp
+++ b/src/hotspot/share/opto/type.cpp
@@ -62,12 +62,14 @@ const Type::TypeInfo Type::_type_info[Type::lastype] = {
   { Bad,             T_ARRAY,      "array:",        false, 
Node::NotAMachineReg, relocInfo::none          },  // Array

 #if defined(PPC64)
+  { Bad,             T_ILLEGAL,    "vectora:",      false, Op_VecA,            
  relocInfo::none          },  // VectorA.
   { Bad,             T_ILLEGAL,    "vectors:",      false, 0,                  
  relocInfo::none          },  // VectorS
   { Bad,             T_ILLEGAL,    "vectord:",      false, Op_RegL,            
  relocInfo::none          },  // VectorD
   { Bad,             T_ILLEGAL,    "vectorx:",      false, Op_VecX,            
  relocInfo::none          },  // VectorX
   { Bad,             T_ILLEGAL,    "vectory:",      false, 0,                  
  relocInfo::none          },  // VectorY
   { Bad,             T_ILLEGAL,    "vectorz:",      false, 0,                  
  relocInfo::none          },  // VectorZ
 #elif defined(S390)
+  { Bad,             T_ILLEGAL,    "vectora:",      false, Op_VecA,            
  relocInfo::none          },  // VectorA.
   { Bad,             T_ILLEGAL,    "vectors:",      false, 0,                  
  relocInfo::none          },  // VectorS
   { Bad,             T_ILLEGAL,    "vectord:",      false, Op_RegL,            
  relocInfo::none          },  // VectorD
   { Bad,             T_ILLEGAL,    "vectorx:",      false, 0,                  
  relocInfo::none          },  // VectorX

Could you please help to have a try?

Thanks,
Ningsheng


> -----Original Message-----
> From: Lindenmaier, Goetz <goetz.lindenma...@sap.com>
> Sent: Monday, September 7, 2020 2:35 PM
> To: Ningsheng Jian <ningsheng.j...@arm.com>; Andrew Dinn <ad...@redhat.com>;
> hotspot-compiler-...@openjdk.java.net; build-dev@openjdk.java.net; Vladimir
> Ivanov <vladimir.x.iva...@oracle.com>; Erik Österlund 
> <erik.osterl...@oracle.com>
> Cc: aarch64-port-...@openjdk.java.net; Doerr, Martin <martin.do...@sap.com>
> Subject: Crashes on ppc/s390 after 8231441: AArch64: Initial SVE backend 
> support
> 
> Hi
> 
> Since that change was pushed, the vm crashes in the build:
> 
> To suppress the following error report, specify this argument # after -XX: or
> in .hotspotrc:  SuppressErrorAt=/type.cpp:1022 # # A fatal error has been 
> detected by
> the Java Runtime Environment:
> #
> #  Internal Error (/usr/work/... /share/opto/type.cpp:1022), pid=28717, 
> tid=28983 #
> assert(_type_info[base()].dual_type != Bad) failed: implement with v-call # # 
> JRE
> version: OpenJDK Runtime Environment (16.0.0.1) (fastdebug build 16.0.0.1-
> internal+0-adhoc.openjdk.jdk)
> # Java VM: OpenJDK 64-Bit Server VM (fastdebug 16.0.0.1-internal+0-
> adhoc.openjdk.jdk, mixed mode, tiered, compressed oops, g1 gc, linux-ppc64) #
> Problematic frame:
> # V  [libjvm.so+0x1bfe22c]  Type::xdual() const+0xfc # # No core dump will be 
> written.
> Core dumps have been disabled. To enable core dumping, try "ulimit -c 
> unlimited"
> before starting Java again
> 
> Do you have an ad-hoc idea of the problem?
> 
> I locally backed out the change which fixed the issue.
> 
> Best regards,
>   Goetz.