netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Gerrit Renker <gerrit@erg.abdn.ac.uk>
To: davem@davemloft.net
Cc: dccp@vger.kernel.org, netdev@vger.kernel.org,
	Gerrit Renker <gerrit@erg.abdn.ac.uk>
Subject: [PATCH 4/5] dccp tfrc/ccid-3: computing the loss rate from the Loss Event Rate
Date: Tue, 21 Sep 2010 12:28:57 +0200	[thread overview]
Message-ID: <1285064938-5933-5-git-send-email-gerrit@erg.abdn.ac.uk> (raw)
In-Reply-To: <1285064938-5933-4-git-send-email-gerrit@erg.abdn.ac.uk>

This adds a function to take care of the following, separate cases occurring in
the computation of the Loss Rate p:

 * 1/(2^32-1) is mapped into 0% as per RFC 4342, 8.5;
 * 1/0        is mapped into 100%, the maximum;
 * to avoid that p = 1/x is rounded down to 0 when x is very large, since this
   means accidentally re-entering slow-start indicated by p == 0, the minimum
   resolution value of p is now returned instead;
 * a bug in ccid3_hc_rx_getsockopt is fixed: 1/0 was mapped into ~0U.

Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
---
 net/dccp/ccids/ccid3.c             |    9 ++++-----
 net/dccp/ccids/lib/tfrc.h          |    1 +
 net/dccp/ccids/lib/tfrc_equation.c |   14 ++++++++++++++
 3 files changed, 19 insertions(+), 5 deletions(-)

--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -400,10 +400,10 @@ static void ccid3_hc_tx_packet_recv(struct sock *sk, struct sk_buff *skb)
 
 	/* Update loss event rate (which is scaled by 1e6) */
 	pinv = opt_recv->ccid3or_loss_event_rate;
-	if (pinv == ~0U || pinv == 0)	       /* see RFC 4342, 8.5   */
+	if (pinv == 0)
 		hc->tx_p = 0;
-	else				       /* can not exceed 100% */
-		hc->tx_p = scaled_div(1, pinv);
+	else
+		hc->tx_p = tfrc_invert_loss_event_rate(pinv);
 
 	/*
 	 * Update allowed sending rate X as per draft rfc3448bis-00, 4.2/3
@@ -834,8 +834,7 @@ static int ccid3_hc_rx_getsockopt(struct sock *sk, const int optname, int len,
 			return -EINVAL;
 		rx_info.tfrcrx_x_recv = hc->rx_x_recv;
 		rx_info.tfrcrx_rtt    = hc->rx_rtt;
-		rx_info.tfrcrx_p      = hc->rx_pinv == 0 ? ~0U :
-					   scaled_div(1, hc->rx_pinv);
+		rx_info.tfrcrx_p      = tfrc_invert_loss_event_rate(hc->rx_pinv);
 		len = sizeof(rx_info);
 		val = &rx_info;
 		break;
--- a/net/dccp/ccids/lib/tfrc.h
+++ b/net/dccp/ccids/lib/tfrc.h
@@ -57,6 +57,7 @@ static inline u32 tfrc_ewma(const u32 avg, const u32 newval, const u8 weight)
 
 extern u32  tfrc_calc_x(u16 s, u32 R, u32 p);
 extern u32  tfrc_calc_x_reverse_lookup(u32 fvalue);
+extern u32  tfrc_invert_loss_event_rate(u32 loss_event_rate);
 
 extern int  tfrc_tx_packet_history_init(void);
 extern void tfrc_tx_packet_history_exit(void);
--- a/net/dccp/ccids/lib/tfrc_equation.c
+++ b/net/dccp/ccids/lib/tfrc_equation.c
@@ -687,3 +687,17 @@ u32 tfrc_calc_x_reverse_lookup(u32 fvalue)
 	index = tfrc_binsearch(fvalue, 0);
 	return (index + 1) * 1000000 / TFRC_CALC_X_ARRSIZE;
 }
+
+/**
+ * tfrc_invert_loss_event_rate  -  Compute p so that 10^6 corresponds to 100%
+ * When @loss_event_rate is large, there is a chance that p is truncated to 0.
+ * To avoid re-entering slow-start in that case, we set p = TFRC_SMALLEST_P > 0.
+ */
+u32 tfrc_invert_loss_event_rate(u32 loss_event_rate)
+{
+	if (loss_event_rate == UINT_MAX)		/* see RFC 4342, 8.5 */
+		return 0;
+	if (unlikely(loss_event_rate == 0))		/* map 1/0 into 100% */
+		return 1000000;
+	return max_t(u32, scaled_div(1, loss_event_rate), TFRC_SMALLEST_P);
+}

  reply	other threads:[~2010-09-21 10:29 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <miscellaneous_dccp_ccid_fixes>
2010-09-21 10:28 ` net-next-2.6 [PATCH 0/5] dccp: miscellaneous CCID fixes and clean-up Gerrit Renker
2010-09-21 10:28   ` [PATCH 1/5] dccp: Add packet type information to CCID-specific option parsing Gerrit Renker
2010-09-21 10:28     ` [PATCH 2/5] dccp: Replace magic CCID-specific numbers by symbolic constants Gerrit Renker
2010-09-21 10:28       ` [PATCH 3/5] dccp ccid-3: remove dead states Gerrit Renker
2010-09-21 10:28         ` Gerrit Renker [this message]
2010-09-21 10:28           ` [PATCH 5/5] dccp ccid-3: Remove redundant 'options_received' struct Gerrit Renker
2010-09-21 23:03   ` net-next-2.6 [PATCH 0/5] dccp: miscellaneous CCID fixes and clean-up David Miller

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=1285064938-5933-5-git-send-email-gerrit@erg.abdn.ac.uk \
    --to=gerrit@erg.abdn.ac.uk \
    --cc=davem@davemloft.net \
    --cc=dccp@vger.kernel.org \
    --cc=netdev@vger.kernel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).