RFR: u212

Andrew John Hughes gnu.andrew at redhat.com
Tue Apr 30 04:04:43 UTC 2019 

Webrev: https://cr.openjdk.java.net/~andrew/aarch32-8/8u212/
Forest: https://icedtea.classpath.org/people/andrew/aarch32 (still
uploading)

I needed an updated AArch32 tree for the upcoming IcedTea 3.12.0
release, so 8u212-b04 was merged and tagged as
jdk8u212-b04-aarch32-190429 (hope that's the right nomenclature)

Changes:
  - S7127191: SA JSDB does not display native symbols correctly for
transported Linux cores
  - S8027434: "-XX:OnOutOfMemoryError" uses fork instead of vfork
  - S8028254: gc/arguments/TestMinInitialErgonomics.java failed with
unexpected initial heap size
  - S8029661: Support TLS v1.2 algorithm in SunPKCS11 provider
  - S8042131: DateTimeFormatterBuilder Mapped-values do not work for
JapaneseDate
  - S8043387: java/time/test/java/util/TestFormatter.java failed.
  - S8044047: Missing null pointer checks for streams
  - S8059038: Create new launcher for SA tools
  - S8065749: [TESTBUG]: gc/arguments/TestG1HeapRegionSize.java fails at
nightly
  - S8068269: RTM tests that assert on non-zero lock statistics are too
strict in RTMTotalCountIncrRate > 1 cases
  - S8076164: [JTextField] When input too long Thai character, cursor's
behavior is odd
  - S8076274: [TESTBUG] Remove @ignore from runtime\NMT\JcmdDetailDiff.java
  - S8076458:
java/util/stream/test/org/openjdk/tests/java/util/stream/FlatMapOpTest.java
timeout
  - S8077608: [TESTBUG] Enable Hotspot jtreg tests to run in agentvm mode
  - S8080932: [TEST_BUG] Test java/awt/BasicStroke/DashStrokeTest.java
fails with Bad script error due to improper @run notation
  - S8129822: Define "headful" jtreg keyword
  - S8132136: [PIT] RTL orientation in JEditorPane is broken
  - S8132985: Crash in freetypescaler.c due to double free
  - S8133108: [PIT] Container size is wrong in JEditorPane
  - S8133731: [TEST_BUG] Unmappable in ASCII character such as Thai
should be escaped in the regtests targeted for a regular non-I18n runs
  - S8133802: replace some <tt> tags (obsolete in html5) in
security-libs docs
  - S8133984: print_compressed_class_space() is only defined in 64-bit VM
  - S8139803: Fix for 8132985 breaks OpenJDK build on windows.
  - S8141491: Unaligned memory access in Bits.c
  - S8145096: Undefined behaviour in HotSpot
  - S8148928: java/util/stream/test/**/SequentialOpTest.java timed out
intermittently
  - S8164656: krb5 does not retry if TCP connection timeouts
  - S8170681: Remove fontconfig header files from JDK source tree
  - S8175120: Remove old tests on kdc timeout policy
  - S8180469: Wrong short form text for supplemental Japanese era
  - S8180904: Hotspot tests running with -agentvm failing due to classpath
  - S8184309: Build warnings from GCC 7.1 on Fedora 26
  - S8185975: PPC64: Fix vsldoi interface according to the ISA
  - S8187364: Unable to enter zero width non-joiner (ZWNJ) symbol in
Swing text component
  - S8189761: COMPANY_NAME, IMPLEMENTOR, BUNDLE_VENDOR, VENDOR, but no
configure flag
  - S8193764: Cannot set COMPANY_NAME when configuring a build
  - S8195153: [test] runtime/6981737/Test6981737.java shouldn't check
'java.vendor' and 'java.vm.vendor' properties
  - S8197429: Increased stack guard causes segfaults on x86-32
  - S8200109: NMT: diff_malloc_site assert(early->flags() ==
current->flags(), "Must be the same memory type")
  - S8200115: System property java.vm.vendor value includes quotation marks
  - S8202088: Japanese new era implementation
  - S8204142: AWT hang occurs when sequenced events arrive out of
sequence in multiple AppContexts
  - S8205432: Replace the placeholder Japanese era name
  - S8206075: On x86, assert on unbound assembler Labels used as branch
targets
  - S8206120: Add test cases for lenient Japanese era parsing
  - S8207070: Webstart app popup on wrong screen in a one-screen setup
changing to multi-monitor
  - S8207152: Placeholder for Japanese new era should be two characters
  - S8207258: Distrust TLS server certificates anchored by Symantec Root CAs
  - S8208480: Test failure: assert(is_bound() || is_unused()) after
JDK-8206075 in C1
  - S8208656: Move java/util/Calendar/CalendarTestScripts tests into OpenJDK
  - S8210633: Cannot parse JapaneseDate string with
DateTimeFormatterBuilder Mapped-values
  - S8210647: libsaproc is being compiled without optimization
  - S8211106: [windows] Update OS detection code to recognize Windows
Server 2019
  - S8211231: BarrierSetC1::generate_referent_check() confuses register
allocator
  - S8211382: ISO2022JP and GB18030 NIO converter issues
  - S8211398: Square character support for the Japanese new era
  - S8211435: Exception in thread "AWT-EventQueue-1"
java.lang.IllegalArgumentException: null source
  - S8211926: Catastrophic size_t underflow in BitMap::*_large methods
  - S8211936: Better String parsing
  - S8212110: Build of saproc.dll broken on Windows 32 bit after JDK-8210647
  - S8212178: Soft reference reclamation race in
com.sun.xml.internal.stream.util.ThreadLocalBufferAllocator
  - S8212914: Test javax/imageio/plugins/bmp/BMP8BPPLoadTest.java fails
  - S8212941: Support new Japanese era in java.time.chrono.JapaneseEra
  - S8213151: [AIX] Some class library files are missing the Classpath
exception
  - S8213154: Update copyright headers of files in src tree that are
missing Classpath exception
  - S8213419: [AArch32] C2 may hang in
MulLNode::Ideal()/MulINode::Ideal() with gcc 8.2.1
  - S8213419: C2 may hang in MulLNode::Ideal()/MulINode::Ideal() with
gcc 8.2.1
  - S8213583: Error while opening the JFileChooser when desktop contains
shortcuts pointing to deleted files.
  - S8213952: Relax DNSName restriction as per RFC 1123
  - S8213983: [macosx] Keyboard shortcut “cmd +`” stops working properly
if popup window is displayed
  - S8213992: Rename and make DieOnSafepointTimeout the diagnostic option
  - S8214059: Undefined behaviour in ADLC
  - S8214061: Buffer written into itself
  - S8214189: [AArch32]
test/hotspot/jtreg/compiler/intrinsics/mathexact/MulExactLConstantTest.java
fails on Windows x64 when run with -XX:-TieredCompilation
  - S8214189:
test/hotspot/jtreg/compiler/intrinsics/mathexact/MulExactLConstantTest.java
fails on Windows x64 when run with -XX:-TieredCompilation
  - S8214206: Fix for JDK-8213419 is broken on 32-bit
  - S8215364: JavaFX crashes on Ubuntu 18.04 with Wayland while using
Swing-FX interop
  - S8215934: G1 Old Gen MemoryPool CollectionUsage.used values don't
reflect mixed GC results
  - S8215976: Fix gmtime_r declaration conflicts in zip.cpp with linux
header files
  - S8216037: Avoid calling vm_update with a NULL name
  - S8216058: [TESTBUG] tools/launcher/VersionCheck.java fails after
JDK-8215992
  - S8216396: Support new Japanese era and new currency code points in
java.lang.Character for Java SE 8
  - S8217305: Missing 0 in java.dll file version cause issues with patch
management software
  - S8217432: MetaspaceGC::_capacity_until_GC exceeds MaxMetaspaceSize
  - S8217520: Remove vm.opt.MaxGCPauseMillis == "null" from
TestOldGenCollectionUsage.java
  - S8217579: TLS_EMPTY_RENEGOTIATION_INFO_SCSV is disabled after 8211883
  - S8217609: New era placeholder not recognized by
java.text.SimpleDateFormat
  - S8217710: Add 5 currency code points to Java SE 8uX
  - S8217753: Enable HotSpot builds on 5.x Linux kernels
  - S8218453: More dynamic RMI interactions
  - S8218613: [TESTBUG] runtime/ErrorHandling tests are building
incorrect testlibrary classes
  - S8218915: Change isJavaIdentifierStart and isJavaIdentifierPart to
handle new code points
  - S8219066: Fuzzing TrueType fonts: setCurrGlyphID()
  - S8219636: Windows build failure after JDK-8207070 8u backport
  - S8219890: Calendar.getDisplayName() returns empty string for new
Japanese Era on some locales
  - S8219961: [ppc64] Increase code size for interpreter generation.
  - S8220397: REGRESSION: JDK-8036003 backport regresses no_strip builds
  - S8220641: [TESTBUG] New test KdcPolicy.java introduced by
JDK-8164656 needs same change as JDK-8190690
  - S8221355: Performance regression after JDK-8155635 backport into 8u

Two AArch32 changes were necessary for the build to complete:

  - S8213419: [AArch32] C2 may hang in
MulLNode::Ideal()/MulINode::Ideal() with gcc 8.2.1
  - S8214189: [AArch32]
test/hotspot/jtreg/compiler/intrinsics/mathexact/MulExactLConstantTest.java
fails on Windows x64 when run with -XX:-TieredCompilation

The first removes the uabs implementation from the aarch32 code as it is
now present in globalDefinitions.hpp (same change was needed for
AArch64). The second converts longs passed to uabs into julongs, as the
method signature was changed as part of the upstream version of 8214189.

Built on AArch32 via the Fedora infrastructure, though not thoroughly
tested as I don't have local access.

Someone else will need to push as I don't have commit rights.

Thanks,
-- 
Andrew :)

Senior Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

PGP Key: ed25519/0xCFDA0F9B35964222 (hkp://keys.gnupg.net)
Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222
https://keybase.io/gnu_andrew

More information about the aarch32-port-dev mailing list