All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jan Kiszka <jan.kiszka@web.de>
To: Manuel Huber <manuel.h87@gmail.com>
Cc: xenomai@xenomai.org
Subject: Re: [Xenomai] rtdm: minimal patch for sys_rtdm_recvmsg
Date: Mon, 17 Jun 2013 07:27:08 +0200	[thread overview]
Message-ID: <51BE9E2C.9090204@web.de> (raw)
In-Reply-To: <51BE2859.8000307@gmail.com>

On 2013-06-16 23:04, Manuel Huber wrote:
> Hello!
> 
> I plan to add a minor change to RTnet which allows to get reception
> timestamps through recvmsg:
> 
> SO_TIMESTAMPNS patch:
> http://sourceforge.net/mailarchive/forum.php?thread_name=51BAB330.70500%40web.de&forum_name=rtnet-developers
> <http://sourceforge.net/mailarchive/forum.php?thread_name=51BAB330.70500%40web.de&forum_name=rtnet-developers>
> 
> 
> Current implementation (of RTDM) in RTAI and Xenomai doesn't fix the
> msg_control and msg_controllen field in struct msghdr. I believe, it
> should be done similar to the linux kernel:
> 
> 1. User specifies how much space has been allocated for control messages
>    - msg_control is set to starting address of buffer
>    - msg_controllen is set to the size of the buffer
> 2. User calls recvmsg
> 3. Control messages are tried to be inserted by some kernel code
> 4. msg_control is set to the next free byte, msg_controllen is updated to
>    the actual space left in the buffer.
> 5. Before copying the struct to the user, msg_control has to be reset to
>    the actual starting address and msg_controllen has to be set to the
>    number of bytes written to the control message buffer rather than the
>    space left.
> 6. Back to user space.
> 
> Current implementation could cause problems: Control messages are
> currently not used but the msg_controllen field inidicates that control
> messages have been written to the buffer; I'm not sure that this is a
> violation of some standard, but at least it's different from the linux
> kernel implementation!
> 
> The patch has been tested on 3.5.7 kernel and is based on SHA
> 69f6cb5ec287afff5ab197528d9eb1177f6de6d6
> (http://git.xenomai.org/xenomai-jki.git) and should work as expected.
> 
> 
> Best Regards
> 
> Manuel
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: Xenomai-0001-rtdm-Fix-msghdr-struct-cmsg-in-sys_rtdm_recvmsg.patch
> Type: text/x-patch
> Size: 1722 bytes
> Desc: not available
> URL:
> <http://www.xenomai.org/pipermail/xenomai/attachments/20130616/18f43ac0/attachment.bin>
> 

Please post patches inline, makes reviewing easier. The explanatory
part, like above, can be put after the "---" so that it won't show up in
the commit log.

> From 79872fff7bf6536007e1a09c353c4194b0def57c Mon Sep 17 00:00:00 2001
> From: Manuel Huber <Manuel.h87@gmail.com>
> Date: Fri, 14 Jun 2013 10:39:50 +0200
> Subject: [PATCH] rtdm: Fix msghdr struct (cmsg) in sys_rtdm_recvmsg
> 
> Whenever a new control message is put into msg_control buffer
> the actual address and the space left is saved to msg_control
> and msg_controllen. This allows adding messages as long as
> there is enough space left in the user-supplied buffer. Both
> fields have to be fixed again before passing them to the user
> by copying the original starting address of the buffer to
> msg_control and saving the actual amount of bytes written to
> the buffer to msg_controllen.
> ---
>  ksrc/skins/rtdm/syscall.c |    7 +++++++
>  1 files changed, 7 insertions(+), 0 deletions(-)
> 
> diff --git a/ksrc/skins/rtdm/syscall.c b/ksrc/skins/rtdm/syscall.c
> index 0ff5d40..e5d9ae0 100644
> --- a/ksrc/skins/rtdm/syscall.c
> +++ b/ksrc/skins/rtdm/syscall.c
> @@ -79,6 +79,7 @@ static int sys_rtdm_recvmsg(struct pt_regs *regs)
>  {
>  	struct task_struct *p = current;
>  	struct msghdr krnl_msg;
> +	void *cmsg_control;
>  	int ret;
>  
>  	if (unlikely(!access_wok(__xn_reg_arg2(regs),
> @@ -88,11 +89,17 @@ static int sys_rtdm_recvmsg(struct pt_regs *regs)
>  					 sizeof(krnl_msg))))
>  		return -EFAULT;
>  
> +	cmsg_control = krnl_msg.msg_control;
> +
>  	ret = __rt_dev_recvmsg(p, __xn_reg_arg1(regs), &krnl_msg,
>  			       __xn_reg_arg3(regs));
>  	if (unlikely(ret < 0))
>  		return ret;
>  
> +	krnl_msg.msg_controllen = (__kernel_size_t)(krnl_msg.msg_control -
> +						    cmsg_control);

So this creates an API between the RTDM core and the driver implementing
msg_control support. Should be documented then (rtdm_recvmsg_handler_t).

> +	krnl_msg.msg_control = cmsg_control;
> +
>  	if (unlikely(__xn_copy_to_user((void __user *)__xn_reg_arg2(regs),
>  				       &krnl_msg, sizeof(krnl_msg))))

Let's just only copy back what can change, msg_flags and msg_controllen,
just like the kernel does.

Jan

>  		return -EFAULT;
> -- 
> 1.7.0.4


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 263 bytes
Desc: OpenPGP digital signature
URL: <http://www.xenomai.org/pipermail/xenomai/attachments/20130617/dd73cdff/attachment.pgp>

  reply	other threads:[~2013-06-17  5:27 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-06-16 21:04 [Xenomai] rtdm: minimal patch for sys_rtdm_recvmsg Manuel Huber
2013-06-17  5:27 ` Jan Kiszka [this message]
2013-06-17  9:53   ` Manuel Huber
2013-06-17 10:04     ` [Xenomai] Fwd: " Manuel Huber
2013-06-22  6:56     ` [Xenomai] " Jan Kiszka
     [not found]       ` <51C5F102.8010803@gmail.com>
2013-06-22 21:35         ` [Xenomai] Problem accessing msg_name in rt_udp_recvmsg Manuel Huber
2013-06-23 12:23           ` Jan Kiszka

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=51BE9E2C.9090204@web.de \
    --to=jan.kiszka@web.de \
    --cc=manuel.h87@gmail.com \
    --cc=xenomai@xenomai.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.