Storing PDF data in README.md breaks Gitlab

Summary

One of our students put a PDF report in READMD.md and Gitlab became unusable for her. Here's an extract from the log...

==> /var/log/gitlab/unicorn/unicorn_stderr.log <==
output error : unknown encoding ASCII-8BIT

==> /var/log/gitlab/gitlab-rails/production.log <==
Completed 500 Internal Server Error in 334ms (ActiveRecord: 23.7ms)

ActionView::Template::Error (invalid byte sequence in UTF-8):
    5:       %strong
    6:         = readme.name
    7:   .file-content.wiki
    8:     = render_readme(readme)
  lib/banzai/filter/emoji_filter.rb:18:in `match'
  lib/banzai/filter/emoji_filter.rb:18:in `match'
  lib/banzai/filter/emoji_filter.rb:18:in `block in call'
  lib/banzai/filter/emoji_filter.rb:14:in `call'
  lib/banzai/pipeline/base_pipeline.rb:21:in `block (2 levels) in singleton class'
  lib/banzai/renderer.rb:117:in `render_result'
  lib/banzai/renderer.rb:146:in `block in cacheless_render'
  lib/gitlab/metrics.rb:88:in `measure'
  lib/banzai/renderer.rb:145:in `cacheless_render'
  lib/banzai/renderer.rb:30:in `render'
  lib/banzai.rb:3:in `render'
  app/helpers/gitlab_markdown_helper.rb:51:in `markdown'
  app/helpers/application_helper.rb:204:in `render_markup'
  app/helpers/tree_helper.rb:16:in `render_readme'
  app/views/projects/tree/_readme.html.haml:8:in `_app_views_projects_tree__readme_html_haml___2496834008807627574_70134792169740'
  app/views/projects/tree/_tree_content.html.haml:31:in `_app_views_projects_tree__tree_content_html_haml__2201343548870009176_70134815445120'
  app/views/projects/_files.html.haml:5:in `_app_views_projects__files_html_haml___1075111661058725741_70134751138760'
  app/views/projects/show.html.haml:94:in `_app_views_projects_show_html_haml___257231136734449135_70134824573740'
  lib/gitlab/request_profiler/middleware.rb:15:in `call'
  lib/gitlab/middleware/go.rb:16:in `call'

Steps to reproduce

It probably depends on the PDF document you use, I can provide one if required.

Expected behavior

Any graceful failure would do.

Actual behavior

500 server error...

snapshot

Relevant logs and/or screenshots

Oops. They're above.

Output of checks

?

Results of GitLab application Check

(For installations with omnibus-gitlab package run and paste the output of: sudo gitlab-rake gitlab:check SANITIZE=true)

(For installations from source run and paste the output of: sudo -u git -H bundle exec rake gitlab:check RAILS_ENV=production SANITIZE=true)

(we will only investigate if the tests are passing)

Results of GitLab environment info

(For installations with omnibus-gitlab package run and paste the output of: sudo gitlab-rake gitlab:env:info)

(For installations from source run and paste the output of: sudo -u git -H bundle exec rake gitlab:env:info RAILS_ENV=production)

Possible fixes

(If you can, link to the line of code that might be responsible for the problem)