This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] gdbserver-base.exp: Copy file to standard output directory in ${board}_download
- From: Pedro Alves <palves at redhat dot com>
- To: Simon Marchi <simon dot marchi at ericsson dot com>, gdb-patches at sourceware dot org
- Cc: qiyaoltc at gmail dot com
- Date: Wed, 13 Apr 2016 12:21:43 +0100
- Subject: Re: [PATCH] gdbserver-base.exp: Copy file to standard output directory in ${board}_download
- Authentication-results: sourceware.org; auth=none
- References: <1460483287-23953-1-git-send-email-simon dot marchi at ericsson dot com> <570D39A7 dot 7040001 at redhat dot com> <570D403F dot 5010406 at ericsson dot com> <570D4262 dot 90201 at redhat dot com> <570D501E dot 1020603 at ericsson dot com>
On 04/12/2016 08:44 PM, Simon Marchi wrote:
> On 16-04-12 02:45 PM, Pedro Alves wrote:
>> On 04/12/2016 07:36 PM, Simon Marchi wrote:
>>> On 16-04-12 02:08 PM, Pedro Alves wrote:
>>>> Don't we need to consider the case of $dest being an absolute path?
>>>
>>> I thought it wouldn't be better not, since we don't even want for tests to
>>> be able to write outside the standard output directory. I guess that if a
>>> test tries to download to /foo/bar, it will end up as
>>> outputs/gdb.blah/thetest/foo/bar. Then I would say the test would be at fault.
>>
>> OK, but I think we should put a comment to the effect in place. After all,
>> we'll be pointing other people that run into this in their boards to this
>> commit. :-)
>>
>> Otherwise LGTM. (Let's wait a bit to give Yao a chance to comment though.)
>
> Good idea. And if that's ok, I'd put the same comment in gdb_remote_download.
>
> What do you think about this?
Fine with me.
Thanks,
Pedro Alves