#15060: outofmem testcase sometimes crashes on i386 -------------------------------------+----------------------------------- Reporter: bgamari | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Runtime System | Version: 8.2.2 Resolution: | Keywords: ci-breakage Operating System: Unknown/Multiple | Architecture: x86 Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+----------------------------------- Comment (by bgamari): `gdb` has the following to say, {{{ outofmem: internal error: getMBlock: mmap: Invalid argument (GHC version 8.5.20180516 for i386_unknown_linux) Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug Program received signal SIGABRT, Aborted. 0xf7d8be2c in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56 56 ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory. (gdb) bt #0 0xf7d8be2c in __GI_raise (sig=sig@entry=6) at ../nptl/sysdeps/unix/sysv/linux/raise.c:56 #1 0xf7d8d4b3 in __GI_abort () at abort.c:89 #2 0x08306486 in rtsFatalInternalErrorFn (s=0x833e2dd "getMBlock: mmap: %s", ap=0xffffb4a4 "^\300\350\367\021") at rts/RtsMessages.c:188 #3 0x08306555 in barf (s=0x833e2dd "getMBlock: mmap: %s") at rts/RtsMessages.c:47 #4 0x08310036 in my_mmap_or_barf (addr=0x77900000, size=size@entry=1074790400, operation=3) at rts/posix/OSMem.c:233 #5 0x0831011b in osGetMBlocks (n=1025) at rts/posix/OSMem.c:291 #6 0x08309eb4 in getCommittedMBlocks (n=1025) at rts/sm/MBlock.c:534 #7 getMBlocks (n=1025) at rts/sm/MBlock.c:580 #8 0x0830a2d0 in alloc_mega_group (node=node@entry=0, mblocks=<optimized out>) at rts/sm/BlockAlloc.c:378 #9 0x0830a70a in allocGroupOnNode (node=0, n=262145) at rts/sm/BlockAlloc.c:405 #10 0x0830d12e in allocateMightFail (cap=0x8407600 <MainCapability>, n=268435459) at rts/sm/Storage.c:884 #11 0x08314a61 in stg_newByteArrayzh () #12 0x08407600 in ?? () Backtrace stopped: previous frame inner to this frame (corrupt stack?) (gdb) }}} -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/15060#comment:2 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler

#15060: outofmem testcase sometimes crashes on i386 -------------------------------------+----------------------------------- Reporter: bgamari | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.4.3 Component: Runtime System | Version: 8.2.2 Resolution: | Keywords: ci-breakage Operating System: Unknown/Multiple | Architecture: x86 Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Phab:D4704 Wiki Page: | -------------------------------------+----------------------------------- Comment (by Ben Gamari ): In [changeset:"34464fed463c7be07d4664e2f4b96eaf1acfc37b/ghc" 34464fed/ghc]: {{{ #!CommitTicketReference repository="ghc" revision="34464fed463c7be07d4664e2f4b96eaf1acfc37b" rts: Don't madvise if mmap failed On 32-bit Linux `outofmem` did not fail with the expected out-of-memory error message, instead failing with, outofmem: internal error: getMBlock: mmap: Invalid argument This happened because, `my_mmap` would attempt to `madvise` even if the `mmap` call failed. So while `mmap` returns `ENOMEM` we nevertheless try to `madvise`, which clobbers `errno`, giving us the unexpected `EINVAL` error. Consequently we don't detect this to be an out-of-memory error. This should fix #15060. Test Plan: `make test TEST=outofmem` on i386 Reviewers: simonmar, erikd Reviewed By: simonmar Subscribers: rwbarton, thomie, carter GHC Trac Issues: #15060 Differential Revision: https://phabricator.haskell.org/D4704 }}} -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/15060#comment:4 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler