Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Peter Jin
The version of scm_rights receiving that I use in my programs will eventually need to support multiple file descriptors, so the buffer is larger. I had totally forgotten about the difference here :) On 7/20/21 9:27 AM, Willy Tarreau wrote: On Tue, Jul 20, 2021 at 03:04:05AM -0500, Peter Jin

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Willy Tarreau
On Tue, Jul 20, 2021 at 03:04:05AM -0500, Peter Jin wrote: > Sorry, after analyzing the code again, it's not a security issue since the > ancillary buffer can only hold one file descriptor. No problem, it's better that way, and thanks for your detailed explanation! Willy

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Willy Tarreau
Hi Peter, first, thanks for bringing this here. On Tue, Jul 20, 2021 at 01:13:58AM -0500, Peter Jin wrote: > 1. The network namespace support seems to be a bit broken. In the function > "my_socketat" (lines 114-129 of src/namespace.c in the latest dev branch), > you attempt to first change

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Willy Tarreau
Hi Lukas, On Tue, Jul 20, 2021 at 08:48:28AM +0200, Lukas Tribus wrote: > Hello, > > > On Tue, 20 Jul 2021 at 08:13, Peter Jin wrote: > > 2. There is a stack buffer overflow found in one of the files. Not > > disclosing it here because this email will end up on the public mailing > > list. If

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Peter Jin
Sorry, after analyzing the code again, it's not a security issue since the ancillary buffer can only hold one file descriptor. Forget everything that I said and focus on # 1 instead. On 7/20/21 1:48 AM, Lukas Tribus wrote: Hello, On Tue, 20 Jul 2021 at 08:13, Peter Jin wrote: 2. There is

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Peter Jin
After analyzing the thing again, I don't think that it's such a big security issue, or even a security issue at all, since it's not remotely exploitable in any way and probably is very specific to how I'm using it (I haven't found anyone else using it in the same way), so here it is:

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Valters Jansons
On Tue, Jul 20, 2021 at 9:49 AM Lukas Tribus wrote: > It's secur...@haproxy.org, it's somehow well hidden in doc/intro.txt > (that is the *starter* guide). > > I would definitely suggest putting it on the website haproxy.org, and > in the repository move it to a different file, like MAINTAINERS.

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Peter Jin
The security issue is not so critical (it's in a seldom-used protocol family that isn't used unless it's explicitly written out in the config file), and to the best of my knowledge (I searched for it on google), I'm the only one who seems to be using it. Should I disclose it here? Also, I'm

Re: HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Lukas Tribus
Hello, On Tue, 20 Jul 2021 at 08:13, Peter Jin wrote: > 2. There is a stack buffer overflow found in one of the files. Not > disclosing it here because this email will end up on the public mailing > list. If there is a "security" email address I could disclose it to, > what is it? It's

HAProxy Network Namespace Support issues, and I also found a security flaw.

2021-07-20 Thread Peter Jin
Hi, I was browsing the code of HAProxy today (I don't use it regularly, but I still like to read the code since it might be useful for my own applications), and I saw three things that concerned me: 1. The network namespace support seems to be a bit broken. In the function "my_socketat"