著者: Mike Tubby 日付: To: exim-users 題目: Re: [exim] Exim 4.94.0.4 works where Exim 4.94 fails with tainted
path 'not permitted'
On 07/11/2020 20:54, Jeremy Harris via Exim-users wrote: > On 07/11/2020 20:43, Mike Tubby via Exim-users wrote:
>> What do I need to know to fix this one?
>
> Either
> - read back through exim-users, which has amply covered tainting
> or
> - start by hauling up the concept index in the docs, and search on
> the string "taint. Then , if that doesn't answer your issue,
> haul the docs sourcefile "spec.xfpt" from git master into an
> editor and do the same (it has been enhanced since the 4.94
> release).
>
Ok, have had a 5 minute scan read ... seems that tainted data is a
{potential} problem, but in my case the variables that I use to build a
path in transport 'local_delivery':
local_delivery:
driver = appendfile
maildir_format = true
directory = /srv/mail/$domain/$local_part/Maildir
create_directory = true
directory_mode = 0770
mode_fail_narrower = false
message_prefix =
message_suffix =
delivery_date_add
envelope_to_add
return_path_add
user = mail
group = mail
mode = 0660
have already been used as keys in a database look-up previously in the
router that cause the transport to be called:
local_delivery:
driver = accept
condition = ${lookup mysql{SELECT
CONCAT(users.username,'@',domains.domain) AS email FROM \
users LEFT JOIN domains ON users.domain_id=domains.id
WHERE \
users.username='${quote_mysql:$local_part}' AND \
domains.domain='${quote_mysql:$domain}' AND \
users.active=1 AND \
domains.active=1}{yes}{no}}
transport = local_delivery
user = mail
group = mail
and have already been sanitised (by accessing a match on the database).
We cannot expect Exim's business logic to know this.
I know that I could write some convoluted SQL to return an untainted
version of the fields but I might as well write: