akkoma/lib/pleroma/web/plugs/uploaded_media.ex

132 lines
3.5 KiB
Elixir
Raw Normal View History

# Pleroma: A lightweight social networking server
# Copyright © 2017-2021 Pleroma Authors <https://pleroma.social/>
# SPDX-License-Identifier: AGPL-3.0-only
2020-06-24 00:03:48 -06:00
defmodule Pleroma.Web.Plugs.UploadedMedia do
2018-11-23 09:40:45 -07:00
@moduledoc """
"""
import Plug.Conn
import Pleroma.Web.Gettext
2018-11-23 09:40:45 -07:00
require Logger
2020-06-14 12:02:57 -06:00
alias Pleroma.Web.MediaProxy
2018-11-23 09:40:45 -07:00
@behaviour Plug
# no slashes
@path "media"
@default_cache_control_header "public, max-age=1209600"
2018-11-23 09:40:45 -07:00
def init(_opts) do
static_plug_opts =
[
headers: %{"cache-control" => @default_cache_control_header},
cache_control_for_etags: @default_cache_control_header
]
2018-11-23 09:40:45 -07:00
|> Keyword.put(:from, "__unconfigured_media_plug")
|> Keyword.put(:at, "/__unconfigured_media_plug")
|> Plug.Static.init()
config = Pleroma.Config.get(Pleroma.Upload)
allowed_mime_types = Keyword.fetch!(config, :allowed_mime_types)
uploader = Keyword.fetch!(config, :uploader)
%{
static_plug_opts: static_plug_opts,
allowed_mime_types: allowed_mime_types,
uploader: uploader
}
2018-11-23 09:40:45 -07:00
end
def call(
%{request_path: <<"/", @path, "/", file::binary>>} = conn,
%{uploader: uploader} = opts
) do
2019-03-12 00:10:19 -06:00
conn =
case fetch_query_params(conn) do
%{query_params: %{"name" => name}} = conn ->
name = escape_header_value(name)
put_resp_header(conn, "content-disposition", ~s[inline; filename="#{name}"])
2019-03-12 00:10:19 -06:00
conn ->
conn
end
2023-05-26 05:06:41 -06:00
|> merge_resp_headers([{"content-security-policy", "sandbox"}])
2019-03-12 00:10:19 -06:00
with {:ok, get_method} <- uploader.get_file(file),
false <- media_is_banned(conn, get_method) do
get_media(conn, get_method, opts)
2018-11-23 09:40:45 -07:00
else
_ ->
conn
|> send_resp(:internal_server_error, dgettext("errors", "Failed"))
2018-11-23 09:40:45 -07:00
|> halt()
end
end
def call(conn, _opts), do: conn
defp media_is_banned(%{request_path: path} = _conn, {:static_dir, _}) do
MediaProxy.in_banned_urls(Pleroma.Upload.base_url() <> path)
2020-06-14 12:02:57 -06:00
end
defp media_is_banned(_, {:url, url}), do: MediaProxy.in_banned_urls(url)
2020-06-14 12:02:57 -06:00
defp media_is_banned(_, _), do: false
2020-06-14 12:02:57 -06:00
Sanitise Content-Type of uploads The lack thereof enables spoofing ActivityPub objects. A malicious user could upload fake activities as attachments and (if having access to remote search) trick local and remote fedi instances into fetching and processing it as a valid object. If uploads are hosted on the same domain as the instance itself, it is possible for anyone with upload access to impersonate(!) other users of the same instance. If uploads are exclusively hosted on a different domain, even the most basic check of domain of the object id and fetch url matching should prevent impersonation. However, it may still be possible to trick servers into accepting bogus users on the upload (sub)domain and bogus notes attributed to such users. Instances which later migrated to a different domain and have a permissive redirect rule in place can still be vulnerable. If — like Akkoma — the fetching server is overly permissive with redirects, impersonation still works. This was possible because Plug.Static also uses our custom MIME type mappings used for actually authentic AP objects. Provided external storage providers don’t somehow return ActivityStream Content-Types on their own, instances using those are also safe against their users being spoofed via uploads. Akkoma instances using the OnlyMedia upload filter cannot be exploited as a vector in this way — IF the fetching server validates the Content-Type of fetched objects (Akkoma itself does this already). However, restricting uploads to only multimedia files may be a bit too heavy-handed. Instead this commit will restrict the returned Content-Type headers for user uploaded files to a safe subset, falling back to generic 'application/octet-stream' for anything else. This will also protect against non-AP payloads as e.g. used in past frontend code injection attacks. It’s a slight regression in user comfort, if say PDFs are uploaded, but this trade-off seems fairly acceptable. (Note, just excluding our own custom types would offer no protection against non-AP payloads and bear a (perhaps small) risk of a silent regression should MIME ever decide to add a canonical extension for ActivityPub objects) Now, one might expect there to be other defence mechanisms besides Content-Type preventing counterfeits from being accepted, like e.g. validation of the queried URL and AP ID matching. Inserting a self-reference into our uploads is hard, but unfortunately *oma does not verify the id in such a way and happily accepts _anything_ from the same domain (without even considering redirects). E.g. Sharkey (and possibly other *keys) seem to attempt to guard against this by immediately refetching the object from its ID, but this is easily circumvented by just uploading two payloads with the ID of one linking to the other. Unfortunately *oma is thus _both_ a vector for spoofing and vulnerable to those spoof payloads, resulting in an easy way to impersonate our users. Similar flaws exists for emoji and media proxy. Subsequent commits will fix this by rigorously sanitising content types in more areas, hardening our checks, improving the default config and discouraging insecure config options.
2024-03-04 09:50:21 -07:00
defp get_safe_mime_type(%{allowed_mime_types: allowed_mime_types} = _opts, mime) do
[maintype | _] = String.split(mime, "/", parts: 2)
if maintype in allowed_mime_types, do: mime, else: "application/octet-stream"
end
defp set_content_type(conn, opts, filepath) do
real_mime = MIME.from_path(filepath)
clean_mime = get_safe_mime_type(opts, real_mime)
put_resp_header(conn, "content-type", clean_mime)
end
defp get_media(conn, {:static_dir, directory}, opts) do
2018-11-23 09:40:45 -07:00
static_opts =
Map.get(opts, :static_plug_opts)
|> Map.put(:at, [@path])
|> Map.put(:from, directory)
Sanitise Content-Type of uploads The lack thereof enables spoofing ActivityPub objects. A malicious user could upload fake activities as attachments and (if having access to remote search) trick local and remote fedi instances into fetching and processing it as a valid object. If uploads are hosted on the same domain as the instance itself, it is possible for anyone with upload access to impersonate(!) other users of the same instance. If uploads are exclusively hosted on a different domain, even the most basic check of domain of the object id and fetch url matching should prevent impersonation. However, it may still be possible to trick servers into accepting bogus users on the upload (sub)domain and bogus notes attributed to such users. Instances which later migrated to a different domain and have a permissive redirect rule in place can still be vulnerable. If — like Akkoma — the fetching server is overly permissive with redirects, impersonation still works. This was possible because Plug.Static also uses our custom MIME type mappings used for actually authentic AP objects. Provided external storage providers don’t somehow return ActivityStream Content-Types on their own, instances using those are also safe against their users being spoofed via uploads. Akkoma instances using the OnlyMedia upload filter cannot be exploited as a vector in this way — IF the fetching server validates the Content-Type of fetched objects (Akkoma itself does this already). However, restricting uploads to only multimedia files may be a bit too heavy-handed. Instead this commit will restrict the returned Content-Type headers for user uploaded files to a safe subset, falling back to generic 'application/octet-stream' for anything else. This will also protect against non-AP payloads as e.g. used in past frontend code injection attacks. It’s a slight regression in user comfort, if say PDFs are uploaded, but this trade-off seems fairly acceptable. (Note, just excluding our own custom types would offer no protection against non-AP payloads and bear a (perhaps small) risk of a silent regression should MIME ever decide to add a canonical extension for ActivityPub objects) Now, one might expect there to be other defence mechanisms besides Content-Type preventing counterfeits from being accepted, like e.g. validation of the queried URL and AP ID matching. Inserting a self-reference into our uploads is hard, but unfortunately *oma does not verify the id in such a way and happily accepts _anything_ from the same domain (without even considering redirects). E.g. Sharkey (and possibly other *keys) seem to attempt to guard against this by immediately refetching the object from its ID, but this is easily circumvented by just uploading two payloads with the ID of one linking to the other. Unfortunately *oma is thus _both_ a vector for spoofing and vulnerable to those spoof payloads, resulting in an easy way to impersonate our users. Similar flaws exists for emoji and media proxy. Subsequent commits will fix this by rigorously sanitising content types in more areas, hardening our checks, improving the default config and discouraging insecure config options.
2024-03-04 09:50:21 -07:00
|> Map.put(:set_content_type, false)
2018-11-23 09:40:45 -07:00
Sanitise Content-Type of uploads The lack thereof enables spoofing ActivityPub objects. A malicious user could upload fake activities as attachments and (if having access to remote search) trick local and remote fedi instances into fetching and processing it as a valid object. If uploads are hosted on the same domain as the instance itself, it is possible for anyone with upload access to impersonate(!) other users of the same instance. If uploads are exclusively hosted on a different domain, even the most basic check of domain of the object id and fetch url matching should prevent impersonation. However, it may still be possible to trick servers into accepting bogus users on the upload (sub)domain and bogus notes attributed to such users. Instances which later migrated to a different domain and have a permissive redirect rule in place can still be vulnerable. If — like Akkoma — the fetching server is overly permissive with redirects, impersonation still works. This was possible because Plug.Static also uses our custom MIME type mappings used for actually authentic AP objects. Provided external storage providers don’t somehow return ActivityStream Content-Types on their own, instances using those are also safe against their users being spoofed via uploads. Akkoma instances using the OnlyMedia upload filter cannot be exploited as a vector in this way — IF the fetching server validates the Content-Type of fetched objects (Akkoma itself does this already). However, restricting uploads to only multimedia files may be a bit too heavy-handed. Instead this commit will restrict the returned Content-Type headers for user uploaded files to a safe subset, falling back to generic 'application/octet-stream' for anything else. This will also protect against non-AP payloads as e.g. used in past frontend code injection attacks. It’s a slight regression in user comfort, if say PDFs are uploaded, but this trade-off seems fairly acceptable. (Note, just excluding our own custom types would offer no protection against non-AP payloads and bear a (perhaps small) risk of a silent regression should MIME ever decide to add a canonical extension for ActivityPub objects) Now, one might expect there to be other defence mechanisms besides Content-Type preventing counterfeits from being accepted, like e.g. validation of the queried URL and AP ID matching. Inserting a self-reference into our uploads is hard, but unfortunately *oma does not verify the id in such a way and happily accepts _anything_ from the same domain (without even considering redirects). E.g. Sharkey (and possibly other *keys) seem to attempt to guard against this by immediately refetching the object from its ID, but this is easily circumvented by just uploading two payloads with the ID of one linking to the other. Unfortunately *oma is thus _both_ a vector for spoofing and vulnerable to those spoof payloads, resulting in an easy way to impersonate our users. Similar flaws exists for emoji and media proxy. Subsequent commits will fix this by rigorously sanitising content types in more areas, hardening our checks, improving the default config and discouraging insecure config options.
2024-03-04 09:50:21 -07:00
conn =
conn
|> set_content_type(opts, conn.request_path)
|> Pleroma.Web.Plugs.StaticNoCT.call(static_opts)
2018-11-23 09:40:45 -07:00
if conn.halted do
conn
else
conn
2019-07-10 04:40:34 -06:00
|> send_resp(:not_found, dgettext("errors", "Not found"))
2018-11-23 09:40:45 -07:00
|> halt()
end
end
defp get_media(conn, {:url, url}, _) do
2018-11-23 09:40:45 -07:00
conn
|> Phoenix.Controller.redirect(external: url)
|> halt()
end
defp get_media(conn, unknown, _) do
2018-11-23 09:40:45 -07:00
Logger.error("#{__MODULE__}: Unknown get startegy: #{inspect(unknown)}")
conn
2019-07-10 04:40:34 -06:00
|> send_resp(:internal_server_error, dgettext("errors", "Internal Error"))
2018-11-23 09:40:45 -07:00
|> halt()
end
defp escape_header_value(value) do
value
|> String.replace("\"", "\\\"")
|> String.replace("\\r", "")
|> String.replace("\\n", "")
end
2018-11-23 09:40:45 -07:00
end