E.g. https://build.chromium.org/p/chromium.fyi/builders/ClangToTLinux/builds/7646/steps/compile/logs/stdio

FAILED: gen/chrome/common/safe_browsing/ipc_protobuf_message_test.pb.h gen/chrome/common/safe_browsing/ipc_protobuf_message_test.pb.cc pyproto/chrome/common/safe_browsing/ipc_protobuf_message_test_pb2.py 
python ../../tools/protoc_wrapper/protoc_wrapper.py ipc_protobuf_message_test.proto --protoc ./protoc --proto-in-dir ../../chrome/common/safe_browsing --cc-out-dir gen/chrome/common/safe_browsing --py-out-dir pyproto/chrome/common/safe_browsing
Protoc has returned non-zero status: -11 .

First bad: 
https://build.chromium.org/p/chromium.fyi/builders/ClangToTLinux/builds/7646/
Clang: r296374

Last good:
https://build.chromium.org/p/chromium.fyi/builders/ClangToTLinux/builds/7645/
Clang: r296355


Changes in that range that are potentially interesting:

------------------------------------------------------------------------
r296355 | ctopper | 2017-02-27 08:15:32 -0800 (Mon, 27 Feb 2017) | 14 lines

[X86] Use APInt instead of SmallBitVector tracking undef elements from getTarget
ConstantBitsFromNode and getConstVector.

Summary:
SmallBitVector uses a malloc for more than 58 bits on a 64-bit target and more t
han 27 bits on a 32-bit target. Some of the vector types we deal with here use m
ore than those number of elements and therefore cause a malloc.

APInt on the other hand supports up to 64 bits without a malloc. That's the maxi
mum number of bits we need here so we can avoid a malloc for all cases by using 
APInt.

Reviewers: RKSimon

Reviewed By: RKSimon

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D30392
------------------------------------------------------------------------

------------------------------------------------------------------------
r296363 | trentxintong | 2017-02-27 10:00:13 -0800 (Mon, 27 Feb 2017) | 11 lines

Fix a bug when unswitching on partial LIV for SwitchInst

Summary: Fix a bug when unswitching on partial LIV for SwitchInst.

Reviewers: hfinkel, efriedma, sanjoy

Reviewed By: sanjoy

Subscribers: david2050, mzolotukhin, llvm-commits

Differential Revision: https://reviews.llvm.org/D29107
------------------------------------------------------------------------

------------------------------------------------------------------------
r296366 | spatel | 2017-02-27 10:13:48 -0800 (Mon, 27 Feb 2017) | 9 lines

[InlineFunction] add nonnull assumptions based on argument attributes

This was suggested in D27855: have the inliner add assumptions, so we don't 
lose nonnull info provided by argument attributes.

This still doesn't solve PR28430 (dyn_cast), but this gets us closer.

https://reviews.llvm.org/D29999

------------------------------------------------------------------------