: Deprecate TFRC_SMALLEST_P
PATCH 7/8 : Deprecate TFRC_SMALLEST_P
[DCCP]: Deprecate TFRC_SMALLEST_P
This patch deprecates the existing use of an arbitrary value TFRC_SMALLEST_P for low-threshold values of p. This avoids masking low-resolution errors. Instead, the code now checks against real boundaries (implemented by preceding patch) and provides warnings whenever a real value falls below the threshold.
If such messages are observed, it is a better solution to take this as an indication that the lookup table needs to be re-engineered.
Changelog:
----------
This patch
* makes handling all TFRC resolution errors local to the TFRC library
* removes unnecessary test whether X_calc is 'infinity' due to p=0 -- this condition is already caught by tfrc_calc_x()
* removes setting ccid3hctx_p = TFRC_SMALLEST_P in ccid3_hc_tx_packet_recv since this is now done by the TFRC library
* updates BUG_ON test in ccid3_hc_tx_no_feedback_timer to take into account that p now is either 0 (and then X_calc is irrelevant), or it is > 0; since the handling of TFRC_SMALLEST_P is now taken care of in the tfrc library
Justification:
--------------
The TFRC code uses a lookup table which has a bounded resolution. The lowest possible value of the loss event rate `p' which can be resolved is currently 0.0001. Substituting this lower threshold for p when p is less than 0.0001 results in a huge, exponentially-growing error. The error can be computed by the following formula:
(f(0.0001) - f(p))/f(p) * 100 for p < 0.0001
Currently the solution is to use an (arbitrary) value
TFRC_SMALLEST_P = 40 * 1E-6 = 0.00004
and to consider all values below this value as `virtually zero'. Due to the exponentially growing resolution error, this is not a good idea, since it hides the fact that the table can not resolve practically occurring cases. Already at p = TFRC_SMALLEST_P, the error is as high as 58.19%!
--- net/dccp/ccids/ccid3.c | 21 +++++---------------- net/dccp/ccids/ccid3.h | 2 -- 2 files changed, 5 insertions(+), 18 deletions(-)
net/dccp/ccids/ccid3.c +5 -16
@@ -137,8 +137,7 @@ static void ccid3_hc_tx_update_x(structstruct ccid3_hc_tx_sock *hctx =ccid3_hc_tx_sk (sk);const __u32 old_x = hctx->ccid3hctx_x; -/* To avoid large error in calcX */ -if (hctx->ccid3hctx_p >=TFRC_SMALLEST_P ) { +if (hctx->ccid3hctx_p >0 ) { hctx->ccid3hctx_x_calc =tfrc_calc_x (hctx->ccid3hctx_s, hctx->ccid3hctx_rtt, hctx->ccid3hctx_p);
@@ -231,12 +230,9 @@ static void ccid3_hc_tx_no_feedback_time* Else * X_recv = X_calc / 4; */ -BUG_ON (hctx->ccid3hctx_p >=TFRC_SMALLEST_P && - hctx->ccid3hctx_x_calc =0 ); +BUG_ON (hctx->ccid3hctx_p && !hctx->ccid3hctx_x_calc); -/* check also if p is zero -> x_calc is infinity? */ -if (hctx->ccid3hctx_p <TFRC_SMALLEST_P || - hctx->ccid3hctx_x_calc >2 * hctx->ccid3hctx_x_recv) +if (hctx->ccid3hctx_x_calc >2 * hctx->ccid3hctx_x_recv) hctx->ccid3hctx_x_recv =max_t (u32, hctx->ccid3hctx_x_recv /2 , hctx->ccid3hctx_s / (2 *TFRC_T_MBI ));else
@@ -449,15 +445,8 @@ static void ccid3_hc_tx_packet_recv(stru pinv = opt_recv->ccid3or_loss_event_rate;if (pinv = ~0U || pinv =0 ) hctx->ccid3hctx_p =0 ; -else { - hctx->ccid3hctx_p =1000000 / pinv; - -if (hctx->ccid3hctx_p <TFRC_SMALLEST_P ) { - hctx->ccid3hctx_p =TFRC_SMALLEST_P ; -ccid3_pr_debug ("%s, sk=%p, Smallest p used!\n" , -dccp_role (sk), sk); - } - } +else + hctx->ccid3hctx_p =1000000 / pinv;dccp_timestamp (sk, &now);
net/dccp/ccids/ccid3.h +0 -2
@@ -51,8 +51,6 @@/* Parameter t_mbi from [RFC 3448, 4.3]: backoff interval in seconds */ #define TFRC_T_MBI 64 -#define TFRC_SMALLEST_P 40 -enum ccid3_options {TFRC_OPT_LOSS_EVENT_RATE =192 ,TFRC_OPT_LOSS_INTERVALS =193 ,
Quoted 3 lines · Gerrit Renker
[DCCP]: Deprecate
TFRC_SMALLEST_PSigned-off-by Gerrit Renker <[email protected]>
Signature · 3 lines
Web: http://wand.net.nz/~iam4 Blog: http://imcdnzl.blogspot.com WAND Network Research Group
Quoting Ian McDonald:
| On 12/2/06, Gerrit Renker <[email protected]> wrote: | > [DCCP]: Deprecate TFRC_SMALLEST_P | > | > Signed-off-by: Gerrit Renker <[email protected]> | | Acked-by: Ian McDonald <[email protected]>
I would like to inform you that I have made a small change which would otherwise cause erratic behaviour:
In ccid3_hc_tx_no_feedback_timer():
/* If (X_calc > 2 * X_recv)
X_recv = max(X_recv/2, s(2*t_mbi)) Else X_recv = X_calc/4
- /* check also if p is zero -> x_calc is infinity? */
- if (hctx->ccid3hctx_p < TFRC_SMALLEST_P ||
- hctx->ccid3hctx_x_calc > 2 * hctx->ccid3hctx_x_recv)
hctx->ccid3hctx_x_recv = max_t(u32, hctx->ccid3hctx_x_recv / 2, hctx->ccid3hctx_s / (2 * TFRC_T_MBI)); else hctx->ccid3hctx_x_recv = hctx->ccid3hctx_x_calc / 4;
==> the previous replacement was:
+ if (hctx->ccid3hctx_x_calc > 2 * hctx->ccid3hctx_x_recv)
==> if p=0, the `else' case would be taken. But X_calc is then either a stale value or (at worst) 0
which would mean that the sender cuts off here. Therefore, I have done the following:
+ if (hctx->ccid3hctx_p = 0 ||
+ hctx->ccid3hctx_x_calc > 2 * hctx->ccid3hctx_x_recv)
==> In this way we avoid using a stale or zero value of X_calc
Keep the Acked-by?
Quoted 3 lines · Gerrit Renker
Quoting Ian McDonald:
| On 12/2/06, Gerrit Renker <[email protected]> wrote:
Keep the Acked-by?
Yes
Signature · 3 lines
Web: http://wand.net.nz/~iam4 Blog: http://imcdnzl.blogspot.com WAND Network Research Group