Adds minimum RTT probing.

Make formatting linux conform

See merge request lukas.prause/tcp-roccet-kernel-module!1

README.md

@@ -1,35 +1,44 @@
 # TCP ROCCET :rocket: (RTT Oriented CUBIC Congestion control ExTension)
 
-TCP ROCCET is a new TCP congestion control
+TCP ROCCET is a new TCP congestion control algorithm that reduces latency by detecting queuing.
-algorithm suited for current cellular 5G NR beyond networks.
+Additionally, it is specially suited for 4G/5G cellular networks.
-It extends the kernel default congestion control CUBIC
-and improves its performance, and additionally solves
-unwanted side effects of CUBIC’s implementation.
-ROCCET uses its own Slow Start, called LAUNCH, where loss
-is not considered as a congestion event.
-The congestion avoidance phase, called ORBITER, uses 
-CUBIC's window growth function and adds, based on RTT
-and ACK rate, congestion events.
 
-NOTE: A paper for TCP ROCCET is currently under review.
+A peer-reviewed [paper on TCP ROCCET](https://opendl.ifip-tc6.org/db/conf/wons/wons2026/1571217211.pdf) was presented at the WONS 2026 conference.
-Paper draft:
-            https://arxiv.org/abs/2510.25281
 
-The oldest Linux kernel version we tested with ROCCET is 6.1.
+## Install with apt
+Add tcp-roccet-dkms to your source lists and install:
+```
+echo "deb [trusted=yes] https://apt.fury.io/timfuchs/ /" | sudo tee /etc/apt/sources.list.d/tcp-roccet.list
+sudo apt update
+sudo apt install tcp-roccet-dkms
+```
 
-## Setup
+### Loading and using the congestion control
+Note: This change is not persistent, the module will not be loaded at boot.
+```
+sudo modprobe tcp_roccet
+sudo sysctl net.ipv4.tcp_congestion_control=roccet
+```
+### Versions
 
+- release-2026-06-12-v1
+    - Adds minimum RTT probing
+
+## Build from source
+
+### Setup
+Kernel headers are required:
 * Debian: `sudo apt install linux-headers-generic`
 * Fedora: `sudo dnf install kernel-devel`
 (`sudo reboot`)
 
-## Build 
+### Build 
 `make`
 
 
 ## Loading & Unloading the Module  
 **1.** Insert into Kernel: `sudo insmod tcp_roccet.ko`
-In case you get the "Invalid module Format" error, it can help reinstalling the kernel-headers.
+In case you get the "Invalid module Format" error, it can help to reinstall the kernel-headers.
 
 **2.** Use the Algorithm:
 * Either via globally loading it: `sudo sysctl net.ipv4.tcp_congestion_control=roccet`
@@ -39,12 +48,12 @@ In case you get the "Invalid module Format" error, it can help reinstalling the
 
 
 ## Debugging (Using kprobe)
-In order to debug the `tcp_roccet` congestion control algorithm, there exists a Kprobe module (`roccet_kprobe.c`). Using this it is possible to inspect events generated by the algorithm. 
+In order to debug the `tcp_roccet` congestion control algorithm, there exists a Kprobe module (`roccet_kprobe.c`). Using this, it is possible to inspect events generated by the algorithm. 
 
-In order to use the Kprobe module the following steps are necessary: 
+In order to use the Kprobe module, the following steps are necessary: 
 
 **1. Specify the event to inspect**
-For this head into the `tcp_roccet.c` source code and find the function the Kprobe should attach to. 
+For this, head into the `tcp_roccet.c` source code and find the function the Kprobe should attach to. 
 
 For Example: 
 ```
@@ -69,7 +78,7 @@ Required files for this are:
 * `tcp_roccet.c` (Optional; Need to remove corresponding entry in Makefile if missing)
 
 **3. Load Kprobe module**
-To then load the Module use 
+To then load the module use 
 `sudo insmod roccet_kprobe.ko` 
 If you are seeing the error "Unknown symbol in module" you need to first load the roccet algorithm.  
 
@@ -78,21 +87,7 @@ To see the trace output use
 `sudo cat /sys/kernel/tracing/trace_pipe`
 
 **5. Unload Kprobe module**
-To remove the module again use 
+To remove the module again, use 
 `sudo rmmod roccet_kprobe` 
 
 _See more info at_ https://docs.kernel.org/trace/kprobes.html and https://www.kernel.org/doc/Documentation/trace/kprobetrace.rst
-
-## Further Info
-* On TCP-CC Ops: 
-    * https://www.yonch.com/tech/linux-tcp-congestion-control-internals
-    * https://docs.ebpf.io/linux/program-type/BPF_PROG_TYPE_STRUCT_OPS/tcp_congestion_ops/
-
-# Setup Development Environment 
-For specific Kernel:  
-1. Download linux source (the version you want to develop for)
-2. Create Config via `make defconfig`
-3. Compile kernel via `make`
-4. Generate clangd Config via `python scripts/clang-tools/gen_compile_commands.py`
-5. Copy `compile_commands.json` to development directory  
-

tcp_roccet.c

@@ -1,4 +1,4 @@
-// SPDX-License-Identifier: GPL-2.0-only
+// SPDX-License-Identifier: GPL-2.0
 /*
  * TCP ROCCET: An RTT-Oriented CUBIC Congestion Control
  * Extension for 5G and Beyond Networks
@@ -14,7 +14,8 @@
  * CUBIC's window growth function and adds, based on RTT
  * and ACK rate, congestion events.
  *
- * A peer-reviewed paper on TCP ROCCET will be presented at the WONS 2026 conference.
+ * A peer-reviewed paper on TCP ROCCET will be presented
+ * at the WONS 2026 conference.
  * A draft of the paper is available here:
  *		https://arxiv.org/abs/2510.25281
  *
@@ -45,12 +46,12 @@
  */
 
 #include "tcp_roccet.h"
-#include "linux/printk.h"
 #include <linux/btf.h>
 #include <linux/btf_ids.h>
 #include <linux/math64.h>
 #include <linux/mm.h>
 #include <linux/module.h>
+#include <linux/moduleparam.h>
 #include <net/tcp.h>
 
 /* Scale factor beta calculation (max_cwnd = snd_cwnd * beta) */
@@ -63,39 +64,18 @@
  */
 #define ROCCET_ALPHA_TIMES_100 20
 
-/* The amount of seconds ROCCET stores a minRTT.
+/* min RTT probe period in seconds */
- * Enable "calculate_min_rtt" first.
+#define ROCCET_NEXT_MIN_RTT_PROBE 5000
- */
-#define ROCCET_RTT_LOOKBACK_S 10
 
 /* Parameters that are specific to the ROCCET-Algorithm */
 static int sr_rtt_upper_bound __read_mostly = 100;
 static int ack_rate_diff_ss __read_mostly = 10;
-static int ack_rate_diff_ca __read_mostly = 200;
-static int calculate_min_rtt __read_mostly = 0;
-static int ignore_loss __read_mostly = 0;
-static int roccet_minRTT_interpolation_factor __read_mostly = 70;
 
 module_param(sr_rtt_upper_bound, int, 0644);
 MODULE_PARM_DESC(sr_rtt_upper_bound, "ROCCET's upper bound for srRTT.");
 module_param(ack_rate_diff_ss, int, 0644);
 MODULE_PARM_DESC(ack_rate_diff_ss,
-		 "ROCCET's threshold to exit slow start if ACK-rate defer by "
+		 "ROCCET's threshold to exit slow start if ACK-rate defer by given amount of segments.");
-		 "given amount of segments.");
-module_param(ack_rate_diff_ca, int, 0644);
-MODULE_PARM_DESC(ack_rate_diff_ca,
-		 "ROCCET's threshold for ack-rate and cum_cwnd, in percantage "
-		 "of the current cwnd.");
-module_param(calculate_min_rtt, int, 0644);
-MODULE_PARM_DESC(calculate_min_rtt,
-		 "Calculate min RTT if no lower RTT occurs after 10 sec.");
-module_param(ignore_loss, int, 0644);
-MODULE_PARM_DESC(ignore_loss, "Ignore loss as a congestion event.");
-module_param(roccet_minRTT_interpolation_factor, int, 0644);
-MODULE_PARM_DESC(
-	roccet_minRTT_interpolation_factor,
-	"ROCCET factor for interpolating the current RTT with the last minRTT "
-	"(minRTT = (factor * currRTT + (100-factor) * minRTT) / 100)");
 
 static int fast_convergence __read_mostly = 1;
 static int beta __read_mostly = 717; /* = 717/1024 (BICTCP_BETA_SCALE) */
@@ -115,99 +95,126 @@ MODULE_PARM_DESC(beta, "beta for multiplicative increase");
 module_param(initial_ssthresh, int, 0644);
 MODULE_PARM_DESC(initial_ssthresh, "initial value of slow start threshold");
 module_param(bic_scale, int, 0444);
-MODULE_PARM_DESC(
+MODULE_PARM_DESC(bic_scale,
-	bic_scale,
 		 "scale (scaled by 1024) value for bic function (bic_scale/1024)");
 module_param(tcp_friendliness, int, 0644);
 MODULE_PARM_DESC(tcp_friendliness, "turn on/off tcp friendliness");
 
-static inline void roccettcp_reset(struct roccettcp *ca)
+static __always_inline void roccettcp_reset(struct roccettcp *ca)
 {
-	memset(ca, 0, offsetof(struct roccettcp, curr_rtt));
+	memset(ca, 0, sizeof(struct roccettcp));
-	ca->bw_limit.sum_cwnd = 1;
+	ca->next_srrtt_check = 0;
-	ca->bw_limit.sum_acked = 1;
+	ca->curr_min_rtt = ~0U;
-	ca->bw_limit.next_check = 0;
+	ca->last_rtt = 0;
-	ca->curr_min_rtt_timed.rtt = ~0U;
+	ca->ece_received = false;
-	ca->curr_min_rtt_timed.time = ~0U;
-	ca->ece_srrtt = 0;
-	ca->ece_cwnd = 2;
 }
 
-static inline void update_min_rtt(struct sock *sk)
+/* Return true if ROCCET is in min RTT probing.
+ */
+static __always_inline bool is_in_min_rtt_probing(struct roccettcp *ca, u32 now)
+{
+	if (ca->probe_min_rtt_until == 0)
+		return false;
+	return before(now, ca->probe_min_rtt_until);
+}
+
+static __always_inline void update_min_rtt(struct sock *sk)
 {
 	struct roccettcp *ca = inet_csk_ca(sk);
-	u32 now = jiffies_to_usecs(tcp_jiffies32);
-
-	if (now - ca->curr_min_rtt_timed.time >
-		    ROCCET_RTT_LOOKBACK_S * USEC_PER_SEC &&
-	    calculate_min_rtt) {
-		u32 new_min_rtt = max(ca->curr_rtt, 1);
-		u32 old_min_rtt = ca->curr_min_rtt_timed.rtt;
-
-		u32 interpolated_min_rtt =
-			(new_min_rtt * roccet_minRTT_interpolation_factor +
-			 old_min_rtt *
-				 (100 - roccet_minRTT_interpolation_factor)) /
-			100;
-
-		ca->curr_min_rtt_timed.rtt = interpolated_min_rtt;
-		ca->curr_min_rtt_timed.time = now;
-	}
 
 	/* Check if new lower min RTT was found. If so, set it directly */
-	if (ca->curr_rtt < ca->curr_min_rtt_timed.rtt) {
+	if (ca->curr_rtt < ca->curr_min_rtt) {
-		ca->curr_min_rtt_timed.rtt = max(ca->curr_rtt, 1);
+		ca->curr_min_rtt = max(ca->curr_rtt, 1);
-		ca->curr_min_rtt_timed.time = now;
+		/* Probe for the min RTT in ROCCET_NEXT_MIN_RTT_PROBE seconds
+		 * if no other update occurs.
+		 */
+		ca->next_min_rtt_probe = jiffies_to_usecs(tcp_jiffies32) +
+					ROCCET_NEXT_MIN_RTT_PROBE * USEC_PER_MSEC;
 	}
 }
 
 /* Return difference between last and current ack rate.
  */
-static inline int get_ack_rate_diff(struct roccettcp *ca)
+static __always_inline int get_ack_rate_diff(struct roccettcp *ca)
 {
-	return ca->ack_rate.last_rate - ca->ack_rate.curr_rate;
+	if (ca->ack_rate.curr_rate < ca->ack_rate.last_rate)
+		return 0;
+	return ca->ack_rate.curr_rate - ca->ack_rate.last_rate;
 }
 
 /* Update ack rate sampled by 100ms.
  */
-static inline void update_ack_rate(struct sock *sk)
+static __always_inline void update_ack_rate(struct sock *sk, u32 acked, u32 now)
 {
 	struct roccettcp *ca = inet_csk_ca(sk);
-	u32 now = jiffies_to_usecs(tcp_jiffies32);
 	u32 interval = USEC_PER_MSEC * 100;
 
-	if ((u32)(now - ca->ack_rate.last_rate_time) >= interval) {
+	if (after(now, ca->ack_rate.last_rate_time + interval)) {
 		ca->ack_rate.last_rate_time = now;
 		ca->ack_rate.last_rate = ca->ack_rate.curr_rate;
 		ca->ack_rate.curr_rate = ca->ack_rate.cnt;
-		ca->ack_rate.cnt = 0;
+		ca->ack_rate.cnt = acked; // start counting for the new interval
 	} else {
-		ca->ack_rate.cnt += 1;
+		ca->ack_rate.cnt += acked;
 	}
 }
 
 /* Compute srRTT.
  */
-static inline void update_srrtt(struct sock *sk)
+static __always_inline void update_srrtt(struct sock *sk)
 {
 	struct roccettcp *ca = inet_csk_ca(sk);
 
-	if (ca->curr_min_rtt_timed.rtt == 0)
+	/* Avoid integer overflow in the calculation below.
-		return;
+	 * This could occur in cases where we have not yet
+	 * received an RTT sample. In these cases, set the
+	 * rtt to a safe value.
+	 */
+	if (ca->curr_rtt < ca->curr_min_rtt) {
+		ca->curr_rtt = max(ca->curr_rtt, 1);
+		ca->curr_min_rtt = ca->curr_rtt;
+	}
+
+	/* Avoid division by zero */
+	if (ca->curr_min_rtt == 0) {
+		ca->curr_min_rtt = max(ca->curr_min_rtt, 1);
+		return; // skip srRTT update
+	}
 
 	/* Calculate the new rRTT (Scaled by 100).
-	 * 100 * ((sRTT - sRTT_min) / sRTT_min)
+	 * 100 * ((sRTT - sRTT_min) / sRTT_min).
+	 *
+	 * curr_min_rtt_timed.rtt is always <= than curr_rtt,
+	 * since this is the minimum of the rtt.
+	 *
+	 * 0 is a valid value for rrtt.
 	 */
-	u32 rRTT = (100 * (ca->curr_rtt - ca->curr_min_rtt_timed.rtt)) /
+	u32 rrtt = (100 * (ca->curr_rtt - ca->curr_min_rtt)) /
-		   ca->curr_min_rtt_timed.rtt;
+		   ca->curr_min_rtt;
 
 	// (1 - alpha) * srRTT + alpha * rRTT
-	ca->curr_srRTT = ((100 - ROCCET_ALPHA_TIMES_100) * ca->curr_srRTT +
+	ca->curr_srrtt = ((100 - ROCCET_ALPHA_TIMES_100) * ca->curr_srrtt +
-			  ROCCET_ALPHA_TIMES_100 * rRTT) /
+			  ROCCET_ALPHA_TIMES_100 * rrtt) / 100;
-			 100;
 }
 
-__bpf_kfunc static void roccettcp_init(struct sock *sk)
+/* Do a ROCCET congestion event.
+ */
+static __always_inline void roccet_congestion_event(struct sock *sk, u32 now)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+	struct roccettcp *ca = inet_csk_ca(sk);
+
+	ca->epoch_start = 0;
+	ca->roccet_last_event_time_us = now;
+	ca->cnt = 100 * tcp_snd_cwnd(tp);
+	ca->last_max_cwnd = tcp_snd_cwnd(tp);
+	tcp_snd_cwnd_set(tp, min(tp->snd_cwnd_clamp,
+					max((tcp_snd_cwnd(tp) * beta) /
+						BICTCP_BETA_SCALE, 2U)));
+	tp->snd_ssthresh = tcp_snd_cwnd(tp);
+}
+
+static void roccettcp_init(struct sock *sk)
 {
 	struct roccettcp *ca = inet_csk_ca(sk);
 
@@ -216,16 +223,15 @@ __bpf_kfunc static void roccettcp_init(struct sock *sk)
 	if (initial_ssthresh)
 		tcp_sk(sk)->snd_ssthresh = initial_ssthresh;
 
-	/* Initial roccet paramters */
+	/* Initial roccet parameters */
 	ca->roccet_last_event_time_us = 0;
-	ca->curr_min_rtt = ~0U;
 	ca->ack_rate.last_rate = 0;
 	ca->ack_rate.last_rate_time = 0;
 	ca->ack_rate.curr_rate = 0;
 	ca->ack_rate.cnt = 0;
 }
 
-__bpf_kfunc static void roccettcp_cwnd_event(struct sock *sk,
+static void roccettcp_cwnd_event(struct sock *sk,
 					     enum tcp_ca_event event)
 {
 	if (event == CA_EVENT_TX_START) {
@@ -295,7 +301,8 @@ static u32 cubic_root(u64 a)
 
 /* Compute congestion window to use.
  */
-static inline void bictcp_update(struct roccettcp *ca, u32 cwnd, u32 acked)
+static __always_inline void bictcp_update(struct roccettcp *ca, u32 cwnd,
+					  u32 acked)
 {
 	u32 delta, bic_target, max_cnt;
 	u64 offs, t;
@@ -340,7 +347,7 @@ static inline void bictcp_update(struct roccettcp *ca, u32 cwnd, u32 acked)
 	 * (so time^3 is done by using 64 bit)
 	 * and without the support of division of 64bit numbers
 	 * (so all divisions are done by using 32 bit)
-	 *  also NOTE the unit of those veriables
+	 *  also NOTE the unit of those variables
 	 *	  time  = (t - K) / 2^bictcp_HZ
 	 *	  c = bic_scale >> 10
 	 * rtt  = (srtt >> 3) / HZ
@@ -368,11 +375,10 @@ static inline void bictcp_update(struct roccettcp *ca, u32 cwnd, u32 acked)
 		bic_target = ca->bic_origin_point + delta;
 
 	/* cubic function - calc bictcp_cnt*/
-	if (bic_target > cwnd) {
+	if (bic_target > cwnd)
 		ca->cnt = cwnd / (bic_target - cwnd);
-	} else {
+	else
 		ca->cnt = 100 * cwnd; /* very small increment*/
-	}
 
 	/* The initial growth of cubic function may be too conservative
 	 * when the available bandwidth is still unknown.
@@ -405,38 +411,45 @@ tcp_friendliness:
 	ca->cnt = max(ca->cnt, 2U);
 }
 
-__bpf_kfunc static void roccettcp_cong_avoid(struct sock *sk, u32 ack,
+static void roccettcp_cong_avoid(struct sock *sk, u32 ack,
 					     u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct roccettcp *ca = inet_csk_ca(sk);
 
 	u32 now = jiffies_to_usecs(tcp_jiffies32);
-	u32 bw_limit_detect = 0;
+	bool evaluate_srrtt = false;
 	u32 roccet_xj;
 	u32 jitter;
-	if (ca->last_rtt > ca->curr_rtt) {
-		jitter = ca->last_rtt - ca->curr_rtt;
-	} else {
-		jitter = ca->curr_rtt - ca->last_rtt;
-	}
 
-	/* Update roccet paramters */
+	if (after(ca->last_rtt, ca->curr_rtt))
-	update_ack_rate(sk);
+		jitter = ca->last_rtt - ca->curr_rtt;
+	else
+		jitter = ca->curr_rtt - ca->last_rtt;
+
+	/* Update roccet parameters */
+	update_ack_rate(sk, acked, now);
 	update_min_rtt(sk);
 	update_srrtt(sk);
 
-	/* Reset ECE handling if we already have more bandwidth
+	/* do not increase the cwnd if in RTT probe */
-	 * than we received the last ECE.
+	if (is_in_min_rtt_probing(ca, now))
+		return;
+
+	/* Probe for min RTT if the last min RTT update is
+	 * ROCCET_NEXT_MIN_RTT_PROBE seconds ago.
 	 */
-	if(ca->ece_srrtt > 0){
+	if (after(now, ca->next_min_rtt_probe)){
-		if(tcp_snd_cwnd(tp) >= ca->ece_cwnd){
+		ca->probe_min_rtt_until = now + min(200, 2 * ca->curr_rtt) * USEC_PER_MSEC;
-			ca->ece_srrtt = 0;
+		ca->cwnd_before_min_rtt_probe = tcp_snd_cwnd(tp);
-		}
+		tcp_snd_cwnd_set(tp, max(tcp_snd_cwnd(tp) / 2, TCP_INIT_CWND));
+		ca->curr_min_rtt = ~0U;
+		return;
 	}
 
 	/* ROCCET drain.
 	 * Do not increase the cwnd for 100ms after a roccet congestion event
+	 * because the buffer is still not drained.
 	 */
 	if (now - ca->roccet_last_event_time_us <= 100 * USEC_PER_MSEC)
 		return;
@@ -445,19 +458,23 @@ __bpf_kfunc static void roccettcp_cong_avoid(struct sock *sk, u32 ack,
 	 * in networks with large buffers of multiple BDP
 	 * Like in cellular networks (5G, ...).
 	 * Or exit LAUNCH if cwnd is too large for application layer
-	 * data rate.
+	 * data rate (tcp cwnd validation).
 	 */
 
-	if ((tcp_in_slow_start(tp) && ca->curr_srRTT > sr_rtt_upper_bound &&
+	if ((tcp_in_slow_start(tp) && ca->curr_srrtt > sr_rtt_upper_bound &&
-	    get_ack_rate_diff(ca) >= ack_rate_diff_ss) ||
+	     get_ack_rate_diff(ca) <= ack_rate_diff_ss) ||
-		(!tcp_is_cwnd_limited(sk) && tcp_in_slow_start(tp))
+	    (!tcp_is_cwnd_limited(sk) && tcp_in_slow_start(tp))) {
-		) {
 		ca->epoch_start = 0;
 
-		/* Handle inital slow start. Here we observe the most problems */
+		/* Handle initial slow start. Here occur most bufferbloat */
 		if (tp->snd_ssthresh == TCP_INFINITE_SSTHRESH) {
 			tcp_sk(sk)->snd_ssthresh = tcp_snd_cwnd(tp) / 2;
-			tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) / 2);
+			/* since this is the initial slow start,
+			 * the min cwnd won't be 1, so the window
+			 * can't be set to 0 by accident.
+			 */
+			tcp_snd_cwnd_set(tp, max(tcp_snd_cwnd(tp) / 2,
+						 TCP_INIT_CWND));
 		} else {
 			tcp_sk(sk)->snd_ssthresh =
 				tcp_snd_cwnd(tp) - (tcp_snd_cwnd(tp) / 3);
@@ -474,61 +491,39 @@ __bpf_kfunc static void roccettcp_cong_avoid(struct sock *sk, u32 ack,
 			return;
 	}
 
-	if (ca->bw_limit.next_check == 0)
+	if (ca->next_srrtt_check == 0)
-		ca->bw_limit.next_check = now + 5 * ca->curr_rtt;
+		ca->next_srrtt_check = now + 5 * ca->curr_rtt;
 
-	ca->bw_limit.sum_cwnd += tcp_snd_cwnd(tp);
+	/* Check if it's time to evaluate the srRTT */
-	ca->bw_limit.sum_acked += acked;
+	if (after(now, ca->next_srrtt_check)) {
-
+		evaluate_srrtt = true;
-	if (ca->bw_limit.next_check < now) {
-		/* We send more data as we got acked in the last 5 RTTs */
-		if ((ca->bw_limit.sum_cwnd * 100) / ca->bw_limit.sum_acked >=
-		    ack_rate_diff_ca)
-			bw_limit_detect = 1;
 
 		/* reset struct and set next end of period */
-		ca->bw_limit.sum_cwnd = 1;
+		ca->next_srrtt_check = now + 5 * ca->curr_rtt;
-
-		/* set to 1 to avoid division by zero */
-		ca->bw_limit.sum_acked = 1;
-		ca->bw_limit.next_check = now + 5 * ca->curr_rtt;
 	}
 
-	/* Respects the jitter of the connection and add it on top of the upper bound
+	/* Respects the jitter of the connection and add it on top of
-	 * for the srRTT
+	 * the upper bound for the srRTT.
 	 */
-	roccet_xj = ((jitter * 100) / ca->curr_min_rtt_timed.rtt) +
+	roccet_xj = ((jitter * 100) / ca->curr_min_rtt) +
 		    sr_rtt_upper_bound;
 	if (roccet_xj < sr_rtt_upper_bound)
 		roccet_xj = sr_rtt_upper_bound;
 
-	/* This is true if we recently received an ECE bit. 
+	/* The srRTT exceeds the upper bound if bufferbloat happens.
-	 * Therefore we should respect the srRTT at this piont.
+	 * Here, we want to reduce the cwnd and drain the buffer.
 	 */
-	if(ca->ece_srrtt < roccet_xj && ca->ece_srrtt > 0)
+	if (ca->curr_srrtt > roccet_xj && evaluate_srrtt) {
-		roccet_xj = ca->ece_srrtt;
+		roccet_congestion_event(sk, now);
-
-	if (ca->curr_srRTT > roccet_xj && (bw_limit_detect || ca->ece_srrtt > 0)) {
-		ca->epoch_start = 0;
-		ca->roccet_last_event_time_us = now;
-		ca->cnt = 100 * tcp_snd_cwnd(tp);
-
-		/* Set Wmax if cwnd is larger than the old Wmax */
-		if (tcp_snd_cwnd(tp) > ca->last_max_cwnd)
-			ca->last_max_cwnd = tcp_snd_cwnd(tp);
-
-		tcp_snd_cwnd_set(tp, min(tp->snd_cwnd_clamp,
-			max((tcp_snd_cwnd(tp) * beta) / BICTCP_BETA_SCALE, 2U)));
-		tp->snd_ssthresh = tcp_snd_cwnd(tp);
 		return;
 	}
 
 	/* Terminates this function if cwnd is not fully utilized.
-	 * In mobile networks like 5G, this termination causes the cwnd to be frozen at
+	 * In mobile networks like 5G, this termination causes the cwnd to be
-	 * an excessively high value. This is because slow start or HyStart massively
+	 * frozen at an excessively high value. This is because slow start or
-	 * exceed the available bandwidth and leave the cwnd at an excessively high
+	 * HyStart massively exceed the available bandwidth and leave the cwnd
-	 * value. The cwnd cannot therefore be fully utilized because it is limited by
+	 * at an excessively high value. The cwnd cannot therefore be fully
-	 * the connection capacity.
+	 * utilized because it is limited by the connection capacity.
 	 */
 	if (!tcp_is_cwnd_limited(sk))
 		return;
@@ -537,15 +532,21 @@ __bpf_kfunc static void roccettcp_cong_avoid(struct sock *sk, u32 ack,
 	tcp_cong_avoid_ai(tp, max(1, ca->cnt), acked);
 }
 
-__bpf_kfunc static u32 roccettcp_recalc_ssthresh(struct sock *sk)
+static u32 roccettcp_recalc_ssthresh(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 	struct roccettcp *ca = inet_csk_ca(sk);
 
-	if (ignore_loss)
+	/* Handle ECN as ROCCET congestion event. */
+	if (ca->ece_received) {
+		ca->ece_received = false;
+		roccet_congestion_event(sk, jiffies_to_usecs(tcp_jiffies32));
 		return tcp_snd_cwnd(tp);
+	}
 
-	/* Don't exit slow start if loss occurs. */
+	/* On loss in slow start enter congestion avoidance
+	 * without a cwnd reduction.
+	 */
 	if (tcp_in_slow_start(tp))
 		return tcp_snd_cwnd(tp);
 
@@ -562,14 +563,15 @@ __bpf_kfunc static u32 roccettcp_recalc_ssthresh(struct sock *sk)
 	return max((tcp_snd_cwnd(tp) * beta) / BICTCP_BETA_SCALE, 2U);
 }
 
-__bpf_kfunc static void roccettcp_state(struct sock *sk, u8 new_state)
+static void roccettcp_state(struct sock *sk, u8 new_state)
 {
 	struct roccettcp *ca = inet_csk_ca(sk);
+
 	if (new_state == TCP_CA_Loss)
 		roccettcp_reset(ca);
 }
 
-__bpf_kfunc static void roccettcp_acked(struct sock *sk,
+static void roccettcp_acked(struct sock *sk,
 					const struct ack_sample *sample)
 {
 	struct roccettcp *ca = inet_csk_ca(sk);
@@ -583,11 +585,12 @@ __bpf_kfunc static void roccettcp_acked(struct sock *sk,
 		return;
 
 	u32 delay = sample->rtt_us;
+
 	if (delay == 0)
 		delay = 1;
 
 	/* first time call or link delay decreases */
-	if (ca->delay_min == 0 || ca->delay_min > delay)
+	if (ca->delay_min == 0 || after(ca->delay_min, delay))
 		ca->delay_min = delay;
 
 	/* Get valid sample for roccet */
@@ -597,18 +600,15 @@ __bpf_kfunc static void roccettcp_acked(struct sock *sk,
 	}
 }
 
-__bpf_kfunc static void roccet_in_ack_event(struct sock *sk, u32 flags)
+static void roccet_in_ack_event(struct sock *sk, u32 flags)
 {
-	struct tcp_sock *tp = tcp_sk(sk);
 	struct roccettcp *ca = inet_csk_ca(sk);
 
 	/* Handle ECE bit.
-	 * Pocessing of ECE events is done in roccettcp_cong_avoid()
+	 * Processing of ECE events is done in roccettcp_recalc_ssthresh()
 	 */
-	if (flags & CA_ACK_ECE) {
+	if (flags & CA_ACK_ECE)
-		ca->ece_srrtt = ca->curr_srRTT;
+		ca->ece_received = true;
-		ca->ece_cwnd = tcp_snd_cwnd(tp);
-	}
 }
 
 static struct tcp_congestion_ops roccet_tcp __read_mostly = {
@@ -624,26 +624,24 @@ static struct tcp_congestion_ops roccet_tcp __read_mostly = {
 	.name = "roccet",
 };
 
-BTF_KFUNCS_START(tcp_roccet_check_kfunc_ids)
-BTF_ID_FLAGS(func, roccettcp_init)
-BTF_ID_FLAGS(func, roccettcp_recalc_ssthresh)
-BTF_ID_FLAGS(func, roccettcp_cong_avoid)
-BTF_ID_FLAGS(func, roccettcp_state)
-BTF_ID_FLAGS(func, roccettcp_cwnd_event)
-BTF_ID_FLAGS(func, roccettcp_acked)
-BTF_KFUNCS_END(tcp_roccet_check_kfunc_ids)
-
-static const struct btf_kfunc_id_set tcp_roccet_kfunc_set = {
-	.owner = THIS_MODULE,
-	.set = &tcp_roccet_check_kfunc_ids,
-};
-
 static int __init roccettcp_register(void)
 {
-	int ret;
-
 	BUILD_BUG_ON(sizeof(struct roccettcp) > ICSK_CA_PRIV_SIZE);
 
+	/*
+	 * Validate parameters to avoid division by zero errors.
+	 */
+	if (beta <= 0 || beta >= BICTCP_BETA_SCALE) {
+		pr_err("roccet: beta must be between 0 and %d\n",
+		       BICTCP_BETA_SCALE);
+		return -EINVAL;
+	}
+
+	if (bic_scale <= 0) {
+		pr_err("roccet: bic_scale must be positive\n");
+		return -EINVAL;
+	}
+
 	/* Precompute a bunch of the scaling factors that are used per-packet
 	 * based on SRTT of 100ms
 	 */
@@ -672,10 +670,6 @@ static int __init roccettcp_register(void)
 	/* divide by bic_scale and by constant Srtt (100ms) */
 	do_div(cube_factor, bic_scale * 10);
 
-	ret = register_btf_kfunc_id_set(BPF_PROG_TYPE_STRUCT_OPS,
-					&tcp_roccet_kfunc_set);
-	if (ret < 0)
-		return ret;
 	return tcp_register_congestion_control(&roccet_tcp);
 }
 
@@ -690,4 +684,3 @@ module_exit(roccettcp_unregister);
 MODULE_AUTHOR("Lukas Prause, Tim Füchsel");
 MODULE_LICENSE("GPL");
 MODULE_DESCRIPTION("ROCCET TCP");
-MODULE_VERSION("1.0");

tcp_roccet.h

@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * TCP ROCCET congestion control interface
  */
@@ -6,25 +7,16 @@
 
 #include <linux/math64.h>
 
-struct AckRate {
+struct ack_rate {
-	u16 last_rate;	/* Last ACK-rate */
+	u32 last_rate;	/* Last ACK-rate */
 	u32 last_rate_time;	/* Timestamp of the last ACK-rate */
-	u16 curr_rate; 	/* Current ACK-rate */
+	u32 curr_rate;	/* Current ACK-rate */
-	u16 cnt; 	/* Used for counting acks */
+	u32 cnt;	/* Used for counting acks */
 };
 
-struct BandwidthLimitDetect {
+/* Based on the BICTCP struct with additions specific
-	u32 sum_cwnd; 	/* sum of cwnd during time interval */
+ * for the ROCCET-Algorithm
-	u32 sum_acked; 	/* sum of received acks during time interval */
+ */
-	u32 next_check; /* end/upper bound of time interval */
-};
-
-struct TimedRTT {
-	u32 time; /* Time of recoding */
-	u32 rtt;  /* Measured RTT */
-};
-
-/* Based on the BICTCP struct with additions specific for the ROCCET-Algorithm */
 struct roccettcp {
 	u32 cnt;		/* increase cwnd by 1 after ACKs */
 	u32 last_max_cwnd;	/* last maximum snd_cwnd */
@@ -32,22 +24,25 @@ struct roccettcp {
 	u32 last_time;		/* time when updated last_cwnd */
 	u32 bic_origin_point;	/* origin point of bic function */
 	u32 bic_K;		/* time to origin point from the
-				beginning of the current epoch */
+				 * beginning of the current epoch
+				 */
 	u32 delay_min;		/* min delay (usec) */
 	u32 epoch_start;	/* beginning of an epoch */
 	u32 ack_cnt;		/* number of acks */
 	u32 tcp_cwnd;		/* estimated tcp cwnd */
 	u32 curr_rtt;		/* the minimum rtt of current round */
 
-	u32 roccet_last_event_time_us; 	/* The last time ROCCET was triggered */
+	u32 roccet_last_event_time_us;	/* The last time ROCCET was
-	u32 ece_cwnd; 	/* The cwnd when a ECE bit was received */
+					 * triggered
-	u32 ece_srrtt; 	/* The srRTT whent the ECE was received */
+					 */
+	bool ece_received;	/* Set to true if an ECE bit was received */
 	u32 curr_min_rtt;	/* The current observed minRTT */
-	struct TimedRTT curr_min_rtt_timed; /* The current observed minRTT with 
+	u32 next_min_rtt_probe; /* Next time to probe the minRTT */
-						the timestamp when it was observed */
+	u32 probe_min_rtt_until;	/* End of minRTT probing period */
-	u32 curr_srRTT; 		/* The srRTT calculated based on the latest ACK */
+	u32 cwnd_before_min_rtt_probe; /* cwnd before min RTT probeing */
-	struct AckRate ack_rate; 	/* The last and the current ACK rate */
+	u32 curr_srrtt;		/* srRTT calculated based on the latest ACK */
-	struct BandwidthLimitDetect bw_limit;
+	u32 next_srrtt_check;	/* Next check for srRTT */
+	struct ack_rate ack_rate;	/* The last and the current ACK rate */
 	u32 last_rtt;			/* Used for jitter calculation */
 };