RFR(S): 8252407: Build failure with gcc-8+ and asan

Kim Barrett kim.barrett at oracle.com
Thu Sep 24 08:59:15 UTC 2020


> On Sep 23, 2020, at 2:10 AM, Eric Liu <eric.c.liu at arm.com> wrote:
> 
> Hi Kim, 
> 
> Sorry for the delay.  
> 
> This patch removes a redundant string copy in NetworkInterface.c to avoid string-truncation
> warning. Other warnings we talked before, which are unable to completely fix in different version
> of gcc, I have to use pragma to suppress them as a workaround. 
> 
> This patch now could compile with gcc-7, gcc-8, gcc-9, gcc-10 both with or without asan.
> 
> [TESTS]
> Jtreg: hotspot::hotspot_all_no_apps, jdk::jdk_core and langtools::tier1.
> No new failure found.
> 
> http://cr.openjdk.java.net/~qfeng/ericliu/jdk/stringop_trunc/webrev.01/
> https://bugs.openjdk.java.net/browse/JDK-8252407

If possible, my preference would be to avoid the pragma cruft and write the
code in such a way that gcc8/9 without asan doesn't warn, and gcc10 doesn't
warn with or without asan. I've kind of lost track in the discussion of all
the variants whether that's actually feasible.




More information about the core-libs-dev mailing list