Able to reproduce the issue on Windows 7, Mac 10.11.5, Ubuntu 14.04 using 51.0.2704.84, canary 53.0.2766.0 as per above steps.

This is regression issue broken in M-42.

Please find below bisect info:
Last good build:42.0.2274.0
First bad build:42.0.2275.0

CHANGELOG URL:
https://chromium.googlesource.com/chromium/src/+log/ab82b75dbda97e7caf83464c1c1665e72554c5b1..640ca0b5920679e4d0e8e8d53251e757d148e540

From above CL, suspecting below:
https://chromium.googlesource.com/chromium/src/+/68ab2c77414eac62cd067a91f7b9b17af3770ae8

raymes@Could you please look into this issue if it is related to your change, else feel free to assign it to an appropriate dev person.

Thanks,

Thanks for the report. I don't think 2 requests will actually hit the server. I believe one of these requests that is seen in the developer tools is as a result of the browser plugin implementation. I ran a local web server and only saw one request being made when loading a PDF from it.

Sam does that seem accurate to you? Closing as WontFix and we can reopen if needed.

Actually I spotted this bug when debugging the JavaEE app, and just ensured myself through chrome debugger. 
The machine I was working on is not reachable for me anymore, so I am not able to send you any logs or system statistics. Maybe this happens on some special condition:

I can repro this locally. I setup a simple server:

  #!/usr/bin/env ruby                                                             
                                                                                
  require 'webrick'                                                               
                                                                                
  root = File.expand_path('.')                                                    
  server = WEBrick::HTTPServer.new(:Port => 8080, :DocumentRoot => root, :RequestCallback => lambda { |req, resp| p req })
                                                                                
  trap 'INT' do server.shutdown end                                               
                                                                                
  server.start


And put the document from the original report in the same directory. When loading the file I can see the webserver printout that the file was requested twice. Interestingly, if I have devtools closed I see three requests. Two with devtools open. The first with devtools closed is a 304 response, and then it requests the document two more times.


Without Devtools:
  GET /Fortran95_notes.pdf HTTP/1.1
  Host: localhost:8080
  Connection: keep-alive
  Cache-Control: max-age=0
  Upgrade-Insecure-Requests: 1
  User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2756.0 Safari/537.36
  Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8
  Accept-Encoding: gzip, deflate, sdch
  Accept-Language: en-US,en;q=0.8
  Range: bytes=1343488-1343488
  If-Range: 1ec0166-1534b4-57600980
  localhost - - [14/Jun/2016:09:47:26 EDT] "GET /Fortran95_notes.pdf HTTP/1.1" 304 0
  http://localhost:8080/ -> /Fortran95_notes.pdf

  GET /Fortran95_notes.pdf HTTP/1.1
  Host: localhost:8080
  Connection: keep-alive
  Cache-Control: max-age=0
  Upgrade-Insecure-Requests: 1
  User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2756.0 Safari/537.36
  Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8
  Accept-Encoding: gzip, deflate, sdch
  Accept-Language: en-US,en;q=0.8
  localhost - - [14/Jun/2016:09:47:26 EDT] "GET /Fortran95_notes.pdf HTTP/1.1" 200 1389748
  http://localhost:8080/ -> /Fortran95_notes.pdf

  GET /Fortran95_notes.pdf HTTP/1.1
  Host: localhost:8080
  Connection: keep-alive
  User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2756.0 Safari/537.36
  Accept: */*
  Referer: http://localhost:8080/Fortran95_notes.pdf
  Accept-Encoding: gzip, deflate, sdch
  Accept-Language: en-US,en;q=0.8
  localhost - - [14/Jun/2016:09:47:26 EDT] "GET /Fortran95_notes.pdf HTTP/1.1" 200 1389748
  http://localhost:8080/Fortran95_notes.pdf -> /Fortran95_notes.pdf


With Devtools:
  GET /Fortran95_notes.pdf HTTP/1.1
  Host: localhost:8080
  Connection: keep-alive
  Pragma: no-cache
  Cache-Control: no-cache
  Upgrade-Insecure-Requests: 1
  User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2756.0 Safari/537.36
  Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8
  Accept-Encoding: gzip, deflate, sdch
  Accept-Language: en-US,en;q=0.8
  localhost - - [14/Jun/2016:09:47:38 EDT] "GET /Fortran95_notes.pdf HTTP/1.1" 200 1389748
  http://localhost:8080/ -> /Fortran95_notes.pdf


  GET /Fortran95_notes.pdf HTTP/1.1
  Host: localhost:8080
  Connection: keep-alive
  Pragma: no-cache
  Cache-Control: no-cache
  User-Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/53.0.2756.0 Safari/537.36
  Accept: */*
  Referer: http://localhost:8080/Fortran95_notes.pdf
  Accept-Encoding: gzip, deflate, sdch
  Accept-Language: en-US,en;q=0.8
  localhost - - [14/Jun/2016:09:47:38 EDT] "GET /Fortran95_notes.pdf HTTP/1.1" 200 1389748
  http://localhost:8080/Fortran95_notes.pdf -> /Fortran95_notes.pdf

Which version of Chrome are you using? With 51 I just see one request, though the output looks very different.

#<WEBrick::HTTPRequest:0x007fbe650b59e8 @config={:ServerName=>"raymes-
...
<other header details>
...
localhost - - [15/Jun/2016:09:51:09 AEST] "GET /Fortran95_notes.pdf HTTP/1.1" 200 1389748
http://localhost:8080/ -> /Fortran95_notes.pdf

I had the same results with Canary.

Isn't this  bug 587709 ?

Looks like it is. dsinclair: please follow up on that bug. If you can reproduce it, it might help us track down the issue.