DMARC-Filter: OpenDMARC Filter v1.3.2 sourceware.org 163903858018
Subject: Random assembler errors after update mingw64-x86_64-binutils-2.35.2-1
To: cygwin AT cygwin DOT com
References: <announce DOT 87o8h52pnz DOT fsf AT Rainer DOT invalid>
Message-ID: <f4d1da5f-897f-0b81-f474-65725c8bf1df@yahoo.de>
Date: Mon, 1 Feb 2021 10:25:57 +0100
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101
 Thunderbird/78.7.0
MIME-Version: 1.0
In-Reply-To: <announce.87o8h52pnz.fsf@Rainer.invalid>
Content-Language: en-US
Precedence: list
From: Michael Soegtrop via Cygwin <cygwin AT cygwin DOT com>
Reply-To: Michael Soegtrop <MSoegtrop AT yahoo DOT de>
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset="us-ascii"; Format="flowed"
Errors-To: cygwin-bounces AT cygwin DOT com
Sender: "Cygwin" <cygwin-bounces AT cygwin DOT com>

Hi Achim,

I have some evidence that this update leads to random assembler errors. 
I have 3 CI failures (out of 5 runs) since the update in an open source 
project I maintain. The failure is always an error from the MinGW 64 
assembler and the errors are random - that is if I rerun the same CI 
test the error went away in one of one tested case and in the failing 
cases the error happened compiling different files.

I haven't seen anything like this before the update.

Here are links to one failing and one good case:

Fail run package version:
https://github.com/coq/platform/runs/1803203762?check_suite_focus=true#step:4:164

Fail run error message:
https://github.com/coq/platform/runs/1803203762?check_suite_focus=true#step:4:1664

Good run package version:
https://github.com/coq/platform/runs/1799374171?check_suite_focus=true#step:4:164

I will revert to the old version of the package and see if the errors go 
away.

The 32 bit version seems fine do far.

Best regards,

Michael
--
Problem reports:      https://cygwin.com/problems.html
FAQ:                  https://cygwin.com/faq/
Documentation:        https://cygwin.com/docs.html
Unsubscribe info:     https://cygwin.com/ml/#unsubscribe-simple