Trace/BPT trap with casperjs on 6.4

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Trace/BPT trap with casperjs on 6.4

cayuga2
After upgrading to 6.4 casperjs seems to be broken.

    % casperjs sample.js
    Trace/BPT trap

This used to work fine with 6.3.  Am I missing something obvious?

Here's my dmesg:

    OpenBSD 6.4 (GENERIC.MP) #364: Thu Oct 11 13:30:23 MDT 2018
    [hidden email]:/usr/src/sys/arch/amd64/compile/GENERIC.MP
    real mem = 1056817152 (1007MB)
    avail mem = 1015574528 (968MB)
    mpath0 at root
    scsibus0 at mpath0: 256 targets
    mainbus0 at root
    bios0 at mainbus0: SMBIOS rev. 2.8 @ 0xf59a0 (9 entries)
    bios0: vendor SeaBIOS version
    "rel-1.11.0-0-g63451fca13-prebuilt.qemu-project.org" date 04/01/2014
    bios0: QEMU Standard PC (i440FX + PIIX, 1996)
    acpi0 at bios0: rev 0
    acpi0: sleep states S3 S4 S5
    acpi0: tables DSDT FACP APIC HPET
    acpi0: wakeup devices
    acpitimer0 at acpi0: 3579545 Hz, 24 bits
    acpimadt0 at acpi0 addr 0xfee00000: PC-AT compat
    cpu0 at mainbus0: apid 0 (boot processor)
    cpu0: Common KVM processor, 2660.44 MHz, 0f-06-01
    cpu0:
    FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,MMX,FXSR,SSE,SSE2,HTT,SSE3,CX16,x2APIC,HV,NXE,LONG,LAHF,MELTDOWN
    cpu0: 64KB 64b/line 2-way I-cache, 64KB 64b/line 2-way D-cache,
    512KB 64b/line 16-way L2 cache
    cpu0: ITLB 255 4KB entries direct-mapped, 255 4MB entries direct-mapped
    cpu0: DTLB 255 4KB entries direct-mapped, 255 4MB entries direct-mapped
    cpu0: smt 0, core 0, package 0
    mtrr: Pentium Pro MTRR support, 8 var ranges, 88 fixed ranges
    cpu0: apic clock running at 1000MHz
    cpu1 at mainbus0: apid 1 (application processor)
    cpu1: Common KVM processor, 2660.10 MHz, 0f-06-01
    cpu1:
    FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,MMX,FXSR,SSE,SSE2,HTT,SSE3,CX16,x2APIC,HV,NXE,LONG,LAHF,MELTDOWN
    cpu1: 64KB 64b/line 2-way I-cache, 64KB 64b/line 2-way D-cache,
    512KB 64b/line 16-way L2 cache
    cpu1: ITLB 255 4KB entries direct-mapped, 255 4MB entries direct-mapped
    cpu1: DTLB 255 4KB entries direct-mapped, 255 4MB entries direct-mapped
    cpu1: smt 0, core 1, package 0
    ioapic0 at mainbus0: apid 0 pa 0xfec00000, version 11, 24 pins
    acpihpet0 at acpi0: 100000000 Hz
    acpiprt0 at acpi0: bus 0 (PCI0)
    acpicpu0 at acpi0: C1(@1 halt!)
    acpicpu1 at acpi0: C1(@1 halt!)
    "ACPI0006" at acpi0 not configured
    acpicmos0 at acpi0
    "PNP0A06" at acpi0 not configured
    "PNP0A06" at acpi0 not configured
    "PNP0A06" at acpi0 not configured
    "QEMU0002" at acpi0 not configured
    "ACPI0010" at acpi0 not configured
    pvbus0 at mainbus0: KVM
    pci0 at mainbus0 bus 0
    pchb0 at pci0 dev 0 function 0 "Intel 82441FX" rev 0x02
    pcib0 at pci0 dev 1 function 0 "Intel 82371SB ISA" rev 0x00
    pciide0 at pci0 dev 1 function 1 "Intel 82371SB IDE" rev 0x00: DMA,
    channel 0 wired to compatibility, channel 1 wired to compatibility
    pciide0: channel 0 disabled (no drives)
    atapiscsi0 at pciide0 channel 1 drive 0
    scsibus1 at atapiscsi0: 2 targets
    cd0 at scsibus1 targ 0 lun 0: <QEMU, QEMU DVD-ROM, 2.5+> ATAPI
    5/cdrom removable
    cd0(pciide0:1:0): using PIO mode 4, DMA mode 2
    uhci0 at pci0 dev 1 function 2 "Intel 82371SB USB" rev 0x01: apic 0
    int 11
    piixpm0 at pci0 dev 1 function 3 "Intel 82371AB Power" rev 0x03:
    apic 0 int 9
    iic0 at piixpm0
    vga1 at pci0 dev 2 function 0 "Bochs VGA" rev 0x02
    wsdisplay0 at vga1 mux 1: console (80x25, vt100 emulation)
    wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
    virtio0 at pci0 dev 3 function 0 "Qumranet Virtio Memory" rev 0x00
    viomb0 at virtio0
    virtio0: apic 0 int 11
    virtio1 at pci0 dev 10 function 0 "Qumranet Virtio Storage" rev 0x00
    vioblk0 at virtio1
    scsibus2 at vioblk0: 2 targets
    sd0 at scsibus2 targ 0 lun 0: <VirtIO, Block Device, > SCSI3
    0/direct fixed
    sd0: 3072MB, 512 bytes/sector, 6291456 sectors
    virtio1: msix shared
    virtio2 at pci0 dev 18 function 0 "Qumranet Virtio Network" rev 0x00
    vio0 at virtio2: address 02:db:a7:f0:8d:52
    virtio2: msix shared
    ppb0 at pci0 dev 30 function 0 "Red Hat Qemu PCI-PCI" rev 0x00
    pci1 at ppb0 bus 1
    ppb1 at pci0 dev 31 function 0 "Red Hat Qemu PCI-PCI" rev 0x00
    pci2 at ppb1 bus 2
    isa0 at pcib0
    isadma0 at isa0
    fdc0 at isa0 port 0x3f0/6 irq 6 drq 2
    pckbc0 at isa0 port 0x60/5 irq 1 irq 12
    pckbd0 at pckbc0 (kbd slot)
    wskbd0 at pckbd0: console keyboard, using wsdisplay0
    pms0 at pckbc0 (aux slot)
    wsmouse0 at pms0 mux 0
    pcppi0 at isa0 port 0x61
    spkr0 at pcppi0
    usb0 at uhci0: USB revision 1.0
    uhub0 at usb0 configuration 1 interface 0 "Intel UHCI root hub" rev
    1.00/1.00 addr 1
    uhidev0 at uhub0 port 1 configuration 1 interface 0 "QEMU QEMU USB
    Tablet" rev 2.00/0.00 addr 2
    uhidev0: iclass 3/0
    ums0 at uhidev0: 3 buttons, Z dir
    wsmouse1 at ums0 mux 0
    vscsi0 at root
    scsibus3 at vscsi0: 256 targets
    softraid0 at root
    scsibus4 at softraid0: 256 targets
    root on sd0a (51d03eeef9af3a89.a) swap on sd0b dump on sd0b
    fd0 at fdc0 drive 1: density unknown



Reply | Threaded
Open this post in threaded view
|

Re: Trace/BPT trap with casperjs on 6.4

Todd Mortimer
On Sat, Oct 27, 2018 at 08:35:45AM -0400, Jeff wrote:
> After upgrading to 6.4 casperjs seems to be broken.
>
>    % casperjs sample.js
>    Trace/BPT trap
>
> This used to work fine with 6.3.  Am I missing something obvious?

This looks an incompatibility between the older webkit engine used in
phantomjs and the retguard security feature introduced in clang in 6.4.
We disabled retguard in the x11/qt5/qtwebkit port for the same reason,
but it looks like nobody noticed that phantomjs was also affected. I am
a bit surprised - retguard was in 6.4-current since last June.

Building the www/phantomjs port with the diff below (generated against
the 6.4-stable ports branch) yields a working phanotmjs (and therefore a
working casper). I imagine the maintainer will have a nicer way to do
it, but this worked for me to get a working phantomjs.


Index: Makefile
===================================================================
RCS file: /cvs/ports/www/phantomjs/Makefile,v
retrieving revision 1.22
diff -u -p -u -r1.22 Makefile
--- Makefile 15 Jul 2018 08:36:39 -0000 1.22
+++ Makefile 27 Oct 2018 23:09:25 -0000
@@ -101,7 +101,7 @@ post-patch:
  ${WRKSRC}/src/qt/qtbase/mkspecs/openbsd-g++/qmake.conf

 MAKE_ENV += OSTYPE=openbsd \
- CC=cc CXX=c++ LINK_C=cc LINK=c++ \
+ CC=cc CXX=c++ LINK_C=cc LINK=c++ CXXFLAGS="${CXXFLAGS} -fno-ret-protector" \
  PYTHON=${MODPY_BIN} RUBY=${RUBY} \
  MAKE=make \
  MAKEFLAGS="-j ${MAKE_JOBS} PYTHON=${MODPY_BIN} RUBY=${RUBY}" \