headless_shell for arm64 crashing in v8/src/codegen/arm64/macro-assembler-arm64.cc UseScratchRegisterScope::AcquuireNextAvailable()

38 views
Skip to first unread message

Timothy Lee

unread,
Mar 22, 2021, 10:16:56 AM3/22/21
to v8-dev
headless_shell: 87.0.4272.0
chromedriver: 87.0.4272.0

Using selenium python and chromedriver to join a Zoom audio/video meeting using headless_shell compiled for arm64. headless_shell crashes at what appears to me to be v8/src/codegen/arm64/macro-assembler-arm64.cc in UseScratchRegisterScope::AcquireNextAvailable(CPURegLIst* available).  It appears to crash always at the point of having joined the meeting and before I start audio or video. From the debug output attached can anyone tell if that is really what is crashing or maybe I'm reading it wrong? The failure appears to be:

# Fatal error in , line 0
# Check failed: !available->IsEmpty().

The only place I found the above check in is the referenced method. Any ideas on what could ultimately be causing the issue?

[0321/225438.515476:VERBOSE1:connection.cc(791)] Conn[3c023520:0:Net[any:0.0.0.x/0:Wildcard:id=0]:tSHjfzkY:1:0:prflx:udp:4.59.156.x:8466->QXdZ3Bla:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: UpdateState(), ms since last received response=745, ms since last received data=1845, rtt=100, pings_since_last_response=
[0321/225438.532021:VERBOSE1:connection.cc(791)] Conn[3c01d610:0:Net[any:0.0.0.x/0:Wildcard:id=0]:5wC2vpwE:1:0:prflx:udp:4.59.156.x:8466->INXF9k6M:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: UpdateState(), ms since last received response=747, ms since last received data=1844, rtt=100, pings_since_last_response=
[0321/225438.766044:VERBOSE1:connection.cc(791)] Conn[3c023520:0:Net[any:0.0.0.x/0:Wildcard:id=0]:tSHjfzkY:1:0:prflx:udp:4.59.156.x:8466->QXdZ3Bla:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: UpdateState(), ms since last received response=996, ms since last received data=2096, rtt=100, pings_since_last_response=
[0321/225438.766435:VERBOSE1:connection.cc(853)] Conn[3c023520:0:Net[any:0.0.0.x/0:Wildcard:id=0]:tSHjfzkY:1:0:prflx:udp:4.59.156.x:8466->QXdZ3Bla:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: Sending STUN ping, id=6d5a77546141325033436c6a, nomination=0
[0321/225438.767092:VERBOSE1:connection.cc(1168)] Conn[3c023520:0:Net[any:0.0.0.x/0:Wildcard:id=0]:tSHjfzkY:1:0:prflx:udp:4.59.156.x:8466->QXdZ3Bla:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: Sent STUN BINDING request, id=6d5a77546141325033436c6a, use_candidate=1, nomination=0
[0321/225438.773630:VERBOSE1:connection.cc(1083)] Conn[3c023520:0:Net[any:0.0.0.x/0:Wildcard:id=0]:tSHjfzkY:1:0:prflx:udp:4.59.156.x:8466->QXdZ3Bla:1:2013266431:local:udp:23.23.115.x:8801|CRWI|S|0|0|7926369428764099582|7]: Received STUN BINDING response, id=6d5a77546141325033436c6a, code=0, rtt=7, pings_since_last_response=6d5a77546141325033436c6a 
[0321/225438.773943:VERBOSE1:connection.cc(398)] Conn[3c023520:0:Net[any:0.0.0.x/0:Wildcard:id=0]:tSHjfzkY:1:0:prflx:udp:4.59.156.x:8466->QXdZ3Bla:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: set_state
[0321/225438.783231:VERBOSE1:connection.cc(791)] Conn[3c01d610:0:Net[any:0.0.0.x/0:Wildcard:id=0]:5wC2vpwE:1:0:prflx:udp:4.59.156.x:8466->INXF9k6M:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: UpdateState(), ms since last received response=998, ms since last received data=2095, rtt=100, pings_since_last_response=
[0321/225438.783572:VERBOSE1:connection.cc(853)] Conn[3c01d610:0:Net[any:0.0.0.x/0:Wildcard:id=0]:5wC2vpwE:1:0:prflx:udp:4.59.156.x:8466->INXF9k6M:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: Sending STUN ping, id=6e384a576f342b61416d7245, nomination=0
[0321/225438.783922:VERBOSE1:connection.cc(1168)] Conn[3c01d610:0:Net[any:0.0.0.x/0:Wildcard:id=0]:5wC2vpwE:1:0:prflx:udp:4.59.156.x:8466->INXF9k6M:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: Sent STUN BINDING request, id=6e384a576f342b61416d7245, use_candidate=1, nomination=0
[0321/225438.790282:VERBOSE1:connection.cc(1083)] Conn[3c01d610:0:Net[any:0.0.0.x/0:Wildcard:id=0]:5wC2vpwE:1:0:prflx:udp:4.59.156.x:8466->INXF9k6M:1:2013266431:local:udp:23.23.115.x:8801|CRWI|S|0|0|7926369428764099582|7]: Received STUN BINDING response, id=6e384a576f342b61416d7245, code=0, rtt=7, pings_since_last_response=6e384a576f342b61416d7245 
[0321/225438.790599:VERBOSE1:connection.cc(398)] Conn[3c01d610:0:Net[any:0.0.0.x/0:Wildcard:id=0]:5wC2vpwE:1:0:prflx:udp:4.59.156.x:8466->INXF9k6M:1:2013266431:local:udp:23.23.115.x:8801|CRWS|S|0|0|7926369428764099582|7]: set_state


#
# Fatal error in , line 0
# Check failed: !available->IsEmpty().
#
#
#
#FailureMessage Object: 0xffff7e630df0#0 0xaaaac88d225c (/opt/chromium/chromium-browser+0x453f25b)
#1 0xaaaac885b87c (/opt/chromium/chromium-browser+0x44c887b)
#2 0xaaaac9da5c14 (/opt/chromium/chromium-browser+0x5a12c13)
#3 0xaaaac9cccf78 (/opt/chromium/chromium-browser+0x5939f77)
#4 0xaaaac822400c (/opt/chromium/chromium-browser+0x3e9100b)
#5 0xaaaac818cee8 (/opt/chromium/chromium-browser+0x3df9ee7)
#6 0xaaaac818d7f0 (/opt/chromium/chromium-browser+0x3dfa7ef)
#7 0xaaaac818c4d4 (/opt/chromium/chromium-browser+0x3df94d3)
#8 0xaaaac8168ad0 (/opt/chromium/chromium-browser+0x3dd5acf)
#9 0xaaaac81568a8 (/opt/chromium/chromium-browser+0x3dc38a7)
#10 0xaaaac815579c (/opt/chromium/chromium-browser+0x3dc279b)
#11 0xaaaac81a3b7c (/opt/chromium/chromium-browser+0x3e10b7b)
#12 0xaaaac81a3904 (/opt/chromium/chromium-browser+0x3e10903)
#13 0xaaaac81c3ff4 (/opt/chromium/chromium-browser+0x3e30ff3)
#14 0xaaaac81c7d64 (/opt/chromium/chromium-browser+0x3e34d63)
#15 0xaaaac9da6574 (/opt/chromium/chromium-browser+0x5a13573)
#16 0xaaaac88aea80 (/opt/chromium/chromium-browser+0x451ba7f)
#17 0xaaaac889adf0 (/opt/chromium/chromium-browser+0x4507def)
#18 0xaaaac88b0bc8 (/opt/chromium/chromium-browser+0x451dbc7)
#19 0xaaaac88b076c (/opt/chromium/chromium-browser+0x451d76b)
#20 0xaaaac88dfaec (/opt/chromium/chromium-browser+0x454caeb)
#21 0xaaaac88b029c (/opt/chromium/chromium-browser+0x451d29b)
#22 0xaaaac88bbe90 (/opt/chromium/chromium-browser+0x4528e8f)
#23 0xaaaac88bbb7c (/opt/chromium/chromium-browser+0x4528b7b)
#24 0xaaaac88e0334 (/opt/chromium/chromium-browser+0x454d333)
#25 0xffff88e3d038 (/lib/libpthread-2.23.so+0x7037)
[0321/225438.971073:VERBOSE1:dns_query.cc(184)] Not supporting parsing a DNS query with multiple questions.
[0321/225438.989159:WARNING:sync_reader.cc(171)] ASR: No room in socket buffer.: Broken pipe (32)
[0321/225438.990758:WARNING:sync_reader.cc(171)] ASR: No room in socket buffer.: Broken pipe (32)


Thank You!

Jakob Kummerow

unread,
Mar 22, 2021, 12:14:13 PM3/22/21
to v8-dev
87.0.4272.0 was some random daily snapshot (from ~half a year ago) and may well be buggy.

Does this still happen with a current version, such as 90.0.4430.0 (M90 Beta) or tip-of-tree (currently: 91.0.4455.0)?


--
--
v8-dev mailing list
v8-...@googlegroups.com
http://groups.google.com/group/v8-dev
---
You received this message because you are subscribed to the Google Groups "v8-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to v8-dev+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/v8-dev/d5d256d4-cf55-4bde-8ba6-42302276a997n%40googlegroups.com.

Timothy Lee

unread,
Mar 22, 2021, 12:50:44 PM3/22/21
to v8-dev
I'm building a newer version to see how it goes, we can't use a beta version or tip version, any thoughts on a good stable version to build? I'm in the middle of a 88.0.4324.182 build but I'm not sure that is any better. I've had a hard time determining what should be a good version to build arm64.

Thanks for the response!

Jakob Kummerow

unread,
Mar 22, 2021, 12:58:39 PM3/22/21
to v8-dev
See omahaproxy.appspot.com for current versions on various channels. Also see https://v8.dev/docs/version-numbers (that's meant for embedding V8 directly, but the core points apply to your case as well).

88.0.4324.182 is a much better basis to start from, because at least it's a Stable build. That said, M88 is outdated; if anyone finds a bug in it, the answer will be "update to M89 and report back". Whether any M89 bugs will still be fixed at this point depends on their severity (and the intrusiveness of the fix).

Timothy Lee

unread,
Mar 22, 2021, 1:23:53 PM3/22/21
to v8-dev
Got ya, thanks for the links and advice, big help. I'll try an M89 if the 88 fails. 
Reply all
Reply to author
Forward
0 new messages