RFR: 8175567: Build of hotspot for arm-vfp-sflt fails (original) (raw)

Bob Vandette bob.vandette at oracle.com
Mon Feb 27 18:15:19 UTC 2017

On Feb 27, 2017, at 1:05 PM, Vladimir Kozlov <vladimir.kozlov at oracle.com> wrote:

That change was done explicitly but that was not tested as I see it: "It guards singlesoftfp() and doublesoftfp() by SOFTFP. This is not used in any openJdk platform. I can not test this on the closed platforms ARM32 and PPC32." May be it should check both? Please, explain if not: #if defined(SOFTFP) || defined(ARM32) You need also update: #endif // SOFTFP

SOFTFP is only use for ARM32 ports so I simplified the ifdef.

In reality, I should have used:

#if defined(ARM32) && !defined(ABI_HARD)

but the only downside is that we end up defining these two functions ARM32 platforms but they are not used in the ABI_HARD case.

Bob.

Thanks, Vladimir On 2/27/17 8:15 AM, Bob Vandette wrote: BUG:

https://bugs.openjdk.java.net/browse/JDK-8175567 DESCRIPTION: The changes that were done under (JDK-8160245 Clean up platform #defines in c1LIR.hpp), broke the arm-vfp-sflt build. Here’s a link to that original bug: https://bugs.openjdk.java.net/browse/JDK-8160245 This change below corrects the issue. diff --git a/src/share/vm/c1/c1LIR.hpp b/src/share/vm/c1/c1LIR.hpp --- a/src/share/vm/c1/c1LIR.hpp +++ b/src/share/vm/c1/c1LIR.hpp @@ -613,7 +613,7 @@ // Platform dependant. static LIROpr doublefpu(int reg1, int reg2 = -1 /fnoreg/); -#ifdef SOFTFP +#ifdef ARM32 static LIROpr singlesoftfp(int reg) { return (LIROpr)(intptrt)((reg << LIROprDesc::reg1shift) | LIROprDesc::floattype | Bob.

More information about the hotspot-dev mailing list