RFR(M): 8038498: Fix includes and C inlining after 8035330
goetz.lindenmaier at sap.com
Fri Mar 28 09:22:11 UTC 2014
I think in this case even a build without precompiled headers
would not have helped. I think the issue arises because different
compilers drop different sets of unused methods from headers
in the debug build.
But if you run with -DDONT_USE_PRECOMPILED_HEADER,
why do you miss headers that often? In the last weeks
there were several such issues!
I pass USE_PRECOMPILED_HEADER=0 to the hotspot build,
but that sets the define above in the build.
From: hotspot-gc-dev [mailto:hotspot-gc-dev-bounces at openjdk.java.net] On Behalf Of Thomas Schatzl
Sent: Freitag, 28. März 2014 09:36
To: David Holmes
Cc: hotspot-gc-dev at openjdk.java.net; hotspot-dev at openjdk.java.net
Subject: Re: RFR(M): 8038498: Fix includes and C inlining after 8035330
On Fri, 2014-03-28 at 09:26 +0100, Thomas Schatzl wrote:
> On Fri, 2014-03-28 at 12:06 +1000, David Holmes wrote:
> > Just as a general comment we should test with precompiled headers
> > disabled as well.
> not sure if it is the same, but e.g. all Solaris builds are already
> built with -DDONT_USE_PRECOMPILED_HEADER in JPRT.
> Without the change the sources build with "make
> USE_PRECOMPILED_HEADERS=0 ..." on linux-x64.
> I am not indicating that there is no problem (and any such problem
> should of course be fixed), but it seems strange that seemingly all but
> the AIX compiler can compile the sources without precompiled headers.
> There may be more changes needed than just adding another build target.
I re-read the issue description from the start and it mentions that the
resulting build does not run due to missing symbols.
More information about the hotspot-dev