summaryrefslogtreecommitdiffstats
path: root/net/dccp
diff options
context:
space:
mode:
authorGerrit Renker <gerrit@erg.abdn.ac.uk>2010-09-14 20:16:59 +0200
committerGerrit Renker <gerrit@erg.abdn.ac.uk>2010-09-15 12:36:01 +0200
commit20cbd3e120a0c20bebe420e1fed0e816730bb988 (patch)
tree1c58cf0b974fd0bb4b5a17846e4711e109901096 /net/dccp
parente1000e: Simplify MSI interrupt testing (diff)
downloadlinux-20cbd3e120a0c20bebe420e1fed0e816730bb988.tar.xz
linux-20cbd3e120a0c20bebe420e1fed0e816730bb988.zip
dccp ccid-3: A lower bound for the inter-packet scheduling algorithm
This fixes a subtle bug in the calculation of the inter-packet gap and shows that t_delta, as it is currently used, is not needed. The algorithm from RFC 5348, 8.3 below continually computes a send time t_nom, which is initialised with the current time t_now; t_gran = 1E6 / HZ specifies the scheduling granularity, s the packet size, and X the sending rate: t_distance = t_nom - t_now; // in microseconds t_delta = min(t_ipi, t_gran) / 2; // `delta' parameter in microseconds if (t_distance >= t_delta) { reschedule after (t_distance / 1000) milliseconds; } else { t_ipi = s / X; // inter-packet interval in usec t_nom += t_ipi; // compute the next send time send packet now; } Problem: -------- Rescheduling requires a conversion into milliseconds (sk_reset_timer()). The highest jiffy resolution with HZ=1000 is 1 millisecond, so using a higher granularity does not make much sense here. As a consequence, values of t_distance < 1000 are truncated to 0. This issue has so far been resolved by using instead if (t_distance >= t_delta + 1000) reschedule after (t_distance / 1000) milliseconds; This is unnecessarily large, a lower bound is t_delta' = max(t_delta, 1000). And it implies a further simplification: a) when HZ >= 500, then t_delta <= t_gran/2 = 10^6/(2*HZ) <= 1000, so that t_delta' = MAX(1000, t_delta) = 1000 (constant value); b) when HZ < 500, then t_delta = 1/2*MIN(rtt, t_ipi, t_gran) <= t_gran/2, so that 1000 <= t_delta' <= t_gran/2. The maximum error of using a constant t_delta in (b) is less than half a jiffy. Fix: ---- The patch replaces t_delta with a constant, whose value depends on CONFIG_HZ, changing the above algorithm to: if (t_distance >= t_delta') reschedule after (t_distance / 1000) milliseconds; where t_delta' = 10^6/(2*HZ) if HZ < 500, and t_delta' = 1000 otherwise. Signed-off-by: Gerrit Renker <gerrit@erg.abdn.ac.uk>
Diffstat (limited to 'net/dccp')
-rw-r--r--net/dccp/ccids/ccid3.c19
-rw-r--r--net/dccp/ccids/ccid3.h18
2 files changed, 21 insertions, 16 deletions
diff --git a/net/dccp/ccids/ccid3.c b/net/dccp/ccids/ccid3.c
index 278e17069322..e9ca0983ac58 100644
--- a/net/dccp/ccids/ccid3.c
+++ b/net/dccp/ccids/ccid3.c
@@ -91,19 +91,16 @@ static inline u64 rfc3390_initial_rate(struct sock *sk)
return scaled_div(w_init << 6, hc->tx_rtt);
}
-/*
- * Recalculate t_ipi and delta (should be called whenever X changes)
+/**
+ * ccid3_update_send_interval - Calculate new t_ipi = s / X_inst
+ * This respects the granularity of X_inst (64 * bytes/second).
*/
static void ccid3_update_send_interval(struct ccid3_hc_tx_sock *hc)
{
- /* Calculate new t_ipi = s / X_inst (X_inst is in 64 * bytes/second) */
hc->tx_t_ipi = scaled_div32(((u64)hc->tx_s) << 6, hc->tx_x);
- /* Calculate new delta by delta = min(t_ipi / 2, t_gran / 2) */
- hc->tx_delta = min_t(u32, hc->tx_t_ipi / 2, TFRC_OPSYS_HALF_TIME_GRAN);
-
- ccid3_pr_debug("t_ipi=%u, delta=%u, s=%u, X=%u\n", hc->tx_t_ipi,
- hc->tx_delta, hc->tx_s, (unsigned)(hc->tx_x >> 6));
+ ccid3_pr_debug("t_ipi=%u, s=%u, X=%u\n", hc->tx_t_ipi,
+ hc->tx_s, (unsigned)(hc->tx_x >> 6));
}
static u32 ccid3_hc_tx_idle_rtt(struct ccid3_hc_tx_sock *hc, ktime_t now)
@@ -332,15 +329,15 @@ static int ccid3_hc_tx_send_packet(struct sock *sk, struct sk_buff *skb)
delay = ktime_us_delta(hc->tx_t_nom, now);
ccid3_pr_debug("delay=%ld\n", (long)delay);
/*
- * Scheduling of packet transmissions [RFC 3448, 4.6]
+ * Scheduling of packet transmissions (RFC 5348, 8.3)
*
* if (t_now > t_nom - delta)
* // send the packet now
* else
* // send the packet in (t_nom - t_now) milliseconds.
*/
- if (delay - (s64)hc->tx_delta >= 1000)
- return (u32)delay / 1000L;
+ if (delay >= TFRC_T_DELTA)
+ return (u32)delay / USEC_PER_MSEC;
ccid3_hc_tx_update_win_count(hc, now);
break;
diff --git a/net/dccp/ccids/ccid3.h b/net/dccp/ccids/ccid3.h
index b7e569c22f36..4a00174a97dd 100644
--- a/net/dccp/ccids/ccid3.h
+++ b/net/dccp/ccids/ccid3.h
@@ -45,12 +45,22 @@
/* Two seconds as per RFC 5348, 4.2 */
#define TFRC_INITIAL_TIMEOUT (2 * USEC_PER_SEC)
-/* In usecs - half the scheduling granularity as per RFC3448 4.6 */
-#define TFRC_OPSYS_HALF_TIME_GRAN (USEC_PER_SEC / (2 * HZ))
-
/* Parameter t_mbi from [RFC 3448, 4.3]: backoff interval in seconds */
#define TFRC_T_MBI 64
+/*
+ * The t_delta parameter (RFC 5348, 8.3): delays of less than %USEC_PER_MSEC are
+ * rounded down to 0, since sk_reset_timer() here uses millisecond granularity.
+ * Hence we can use a constant t_delta = %USEC_PER_MSEC when HZ >= 500. A coarse
+ * resolution of HZ < 500 means that the error is below one timer tick (t_gran)
+ * when using the constant t_delta = t_gran / 2 = %USEC_PER_SEC / (2 * HZ).
+ */
+#if (HZ >= 500)
+# define TFRC_T_DELTA USEC_PER_MSEC
+#else
+# define TFRC_T_DELTA (USEC_PER_SEC / (2 * HZ))
+#endif
+
enum ccid3_options {
TFRC_OPT_LOSS_EVENT_RATE = 192,
TFRC_OPT_LOSS_INTERVALS = 193,
@@ -90,7 +100,6 @@ enum ccid3_hc_tx_states {
* @tx_no_feedback_timer: Handle to no feedback timer
* @tx_t_ld: Time last doubled during slow start
* @tx_t_nom: Nominal send time of next packet
- * @tx_delta: Send timer delta (RFC 3448, 4.6) in usecs
* @tx_hist: Packet history
* @tx_options_received: Parsed set of retrieved options
*/
@@ -109,7 +118,6 @@ struct ccid3_hc_tx_sock {
struct timer_list tx_no_feedback_timer;
ktime_t tx_t_ld;
ktime_t tx_t_nom;
- u32 tx_delta;
struct tfrc_tx_hist_entry *tx_hist;
struct ccid3_options_received tx_options_received;
};