* [PATCH v4 4/6] net/gve: add periodic NIC clock synchronization
From: Mark Blasko @ 2026-06-13 4:22 UTC (permalink / raw)
To: stephen; +Cc: dev, joshwash, jtranoleary, blasko
In-Reply-To: <20260613042300.3760470-1-blasko@google.com>
Introduce a mechanism to periodically fetch the NIC hardware timestamp
using the GVE_ADMINQ_REPORT_NIC_TIMESTAMP AdminQ command. The
synchronization runs every 250ms using rte_alarm. If the read fails,
the alarm is still rescheduled. After 7 consecutive failures, the
timestamp is marked as stale, indicating to the RX path that
reconstructed timestamps may be unreliable.
Atomics exist because of the potential for async callers (introduced
here) and async callers (introduced later in the RX datapath) accessing
the cached state.
Signed-off-by: Mark Blasko <blasko@google.com>
Reviewed-by: Joshua Washington <joshwash@google.com>
Reviewed-by: Jasper Tran O'Leary <jtranoleary@google.com>
---
v4:
- Replace EAL alarm thread with a dedicated control thread to
prevent blocking shared EAL interrupt thread for up to 2s
on GVE AdminQ timeouts.
- Implement incremental sleep loop to support fast termination.
v2:
- Removed redundant void* casts.
- Handled alarm reschedule failures by marking timestamp stale.
- Added transient error logging on memzone allocation failure.
---
drivers/net/gve/gve_ethdev.c | 122 +++++++++++++++++++++++++++++++++++
drivers/net/gve/gve_ethdev.h | 12 ++++
2 files changed, 134 insertions(+)
diff --git a/drivers/net/gve/gve_ethdev.c b/drivers/net/gve/gve_ethdev.c
index 476b2c311f..fa91575b3e 100644
--- a/drivers/net/gve/gve_ethdev.c
+++ b/drivers/net/gve/gve_ethdev.c
@@ -452,6 +452,94 @@ gve_dev_start(struct rte_eth_dev *dev)
return 0;
}
+static void
+gve_read_nic_clock(void *arg)
+{
+ struct gve_priv *priv = arg;
+ uint32_t fails;
+ uint64_t ts;
+ int err;
+
+ if (!priv || !priv->nic_ts_report_mz)
+ return;
+
+ memset(priv->nic_ts_report, 0, sizeof(struct gve_nic_ts_report));
+
+ err = gve_adminq_report_nic_timestamp(priv, priv->nic_ts_report_mz->iova);
+ if (err == 0) {
+ ts = be64_to_cpu(priv->nic_ts_report->nic_timestamp);
+ rte_atomic_store_explicit(&priv->last_read_nic_timestamp, ts,
+ rte_memory_order_relaxed);
+ PMD_DRV_LOG(DEBUG, "Fetched NIC Timestamp: %" PRIu64, ts);
+ rte_atomic_store_explicit(&priv->nic_ts_read_fails, 0,
+ rte_memory_order_relaxed);
+ rte_atomic_store_explicit(&priv->nic_ts_stale, 0,
+ rte_memory_order_release);
+ } else {
+ PMD_DRV_LOG(ERR, "Failed to read NIC clock, AQ err: %d", err);
+ fails = rte_atomic_fetch_add_explicit(&priv->nic_ts_read_fails, 1,
+ rte_memory_order_relaxed) + 1;
+ if (fails >= GVE_NIC_CLOCK_READ_MAX_FAILS) {
+ if (!rte_atomic_load_explicit(&priv->nic_ts_stale,
+ rte_memory_order_relaxed))
+ PMD_DRV_LOG(ERR,
+ "NIC timestamping marked as stale after %u consecutive failures",
+ GVE_NIC_CLOCK_READ_MAX_FAILS);
+ rte_atomic_store_explicit(&priv->nic_ts_stale, 1,
+ rte_memory_order_release);
+ }
+ }
+}
+
+static uint32_t
+gve_nic_ts_thread(void *arg)
+{
+ struct gve_priv *priv = arg;
+ unsigned int sleep_cnt;
+
+ while (!rte_atomic_load_explicit(&priv->nic_ts_thread_should_stop,
+ rte_memory_order_relaxed)) {
+ gve_read_nic_clock(priv);
+ for (sleep_cnt = 0; sleep_cnt < GVE_NIC_CLOCK_READ_PERIOD_MS / 10; sleep_cnt++) {
+ if (rte_atomic_load_explicit(&priv->nic_ts_thread_should_stop,
+ rte_memory_order_relaxed))
+ break;
+ rte_delay_us_sleep(10000);
+ }
+ }
+ return 0;
+}
+
+static int
+gve_alloc_nic_ts_report(struct gve_priv *priv)
+{
+ char z_name[RTE_MEMZONE_NAMESIZE];
+
+ snprintf(z_name, sizeof(z_name), "gve_%s_nic_ts_report",
+ priv->pci_dev->device.name);
+ priv->nic_ts_report_mz = rte_memzone_reserve_aligned(z_name,
+ sizeof(struct gve_nic_ts_report), rte_socket_id(),
+ RTE_MEMZONE_IOVA_CONTIG, PAGE_SIZE);
+
+ if (!priv->nic_ts_report_mz) {
+ PMD_DRV_LOG(ERR, "Failed to allocate memzone for NIC TS report");
+ return -ENOMEM;
+ }
+ priv->nic_ts_report = priv->nic_ts_report_mz->addr;
+ rte_atomic_store_explicit(&priv->nic_ts_read_fails, 0, rte_memory_order_relaxed);
+ return 0;
+}
+
+static void
+gve_free_nic_ts_report(struct gve_priv *priv)
+{
+ if (priv->nic_ts_report_mz) {
+ rte_memzone_free(priv->nic_ts_report_mz);
+ priv->nic_ts_report_mz = NULL;
+ priv->nic_ts_report = NULL;
+ }
+}
+
static int
gve_dev_stop(struct rte_eth_dev *dev)
{
@@ -586,6 +674,13 @@ gve_teardown_device_resources(struct gve_priv *priv)
err);
}
+ if (priv->nic_ts_report_mz) {
+ rte_atomic_store_explicit(&priv->nic_ts_thread_should_stop, 1,
+ rte_memory_order_relaxed);
+ rte_thread_join(priv->nic_ts_thread_id, NULL);
+ gve_free_nic_ts_report(priv);
+ }
+
gve_free_ptype_lut_dqo(priv);
gve_free_counter_array(priv);
gve_free_irq_db(priv);
@@ -1252,6 +1347,32 @@ pci_dev_msix_vec_count(struct rte_pci_device *pdev)
return 0;
}
+static void
+gve_setup_nic_timestamp(struct gve_priv *priv)
+{
+ int err;
+
+ if (!priv->nic_timestamp_supported)
+ return;
+
+ rte_atomic_store_explicit(&priv->nic_ts_read_fails, 0, rte_memory_order_relaxed);
+ rte_atomic_store_explicit(&priv->nic_ts_stale, 1, rte_memory_order_relaxed);
+ err = gve_alloc_nic_ts_report(priv);
+ if (err == 0) {
+ gve_read_nic_clock(priv);
+ rte_atomic_store_explicit(&priv->nic_ts_thread_should_stop, 0,
+ rte_memory_order_relaxed);
+ err = rte_thread_create_internal_control(&priv->nic_ts_thread_id, "gve-ts",
+ gve_nic_ts_thread, priv);
+ if (err != 0) {
+ PMD_DRV_LOG(ERR, "Failed to create NIC clock sync thread, err=%d", err);
+ gve_free_nic_ts_report(priv);
+ }
+ } else {
+ PMD_DRV_LOG(ERR, "Failed to allocate memory for NIC timestamping subsystem. Please reset device to retry.");
+ }
+}
+
static int
gve_setup_device_resources(struct gve_priv *priv)
{
@@ -1307,6 +1428,7 @@ gve_setup_device_resources(struct gve_priv *priv)
goto free_ptype_lut;
}
}
+ gve_setup_nic_timestamp(priv);
gve_set_device_resources_ok(priv);
diff --git a/drivers/net/gve/gve_ethdev.h b/drivers/net/gve/gve_ethdev.h
index b67f82c263..4dcbaa9971 100644
--- a/drivers/net/gve/gve_ethdev.h
+++ b/drivers/net/gve/gve_ethdev.h
@@ -12,6 +12,8 @@
#include <rte_pci.h>
#include <pthread.h>
#include <rte_bitmap.h>
+#include <rte_memzone.h>
+#include <rte_thread.h>
#include "base/gve.h"
@@ -39,6 +41,9 @@
#define GVE_RSS_HASH_KEY_SIZE 40
#define GVE_RSS_INDIR_SIZE 128
+#define GVE_NIC_CLOCK_READ_PERIOD_MS 250
+#define GVE_NIC_CLOCK_READ_MAX_FAILS 7
+
#define GVE_TX_CKSUM_OFFLOAD_MASK ( \
RTE_MBUF_F_TX_L4_MASK | \
RTE_MBUF_F_TX_TCP_SEG)
@@ -359,6 +364,13 @@ struct gve_priv {
/* HW Timestamping Fields */
bool nic_timestamp_supported;
+ const struct rte_memzone *nic_ts_report_mz;
+ struct gve_nic_ts_report *nic_ts_report;
+ RTE_ATOMIC(uint64_t) last_read_nic_timestamp;
+ RTE_ATOMIC(uint32_t) nic_ts_read_fails;
+ RTE_ATOMIC(uint8_t) nic_ts_stale;
+ rte_thread_t nic_ts_thread_id;
+ RTE_ATOMIC(uint8_t) nic_ts_thread_should_stop;
};
static inline bool
--
2.54.0.1136.gdb2ca164c4-goog
^ permalink raw reply related
* [PATCH v4 3/6] net/gve: add AdminQ command for NIC timestamps
From: Mark Blasko @ 2026-06-13 4:22 UTC (permalink / raw)
To: stephen; +Cc: dev, joshwash, jtranoleary, blasko
In-Reply-To: <20260613042300.3760470-1-blasko@google.com>
Introduce the necessary definitions and functions for the
GVE_ADMINQ_REPORT_NIC_TIMESTAMP AdminQ command.
Signed-off-by: Mark Blasko <blasko@google.com>
Reviewed-by: Joshua Washington <joshwash@google.com>
Reviewed-by: Jasper Tran O'Leary <jtranoleary@google.com>
---
v2:
- Added adminq timestamp counter reset to gve_adminq_alloc.
---
drivers/net/gve/base/gve_adminq.c | 20 ++++++++++++++++++++
drivers/net/gve/base/gve_adminq.h | 20 ++++++++++++++++++++
drivers/net/gve/gve_ethdev.h | 1 +
3 files changed, 41 insertions(+)
diff --git a/drivers/net/gve/base/gve_adminq.c b/drivers/net/gve/base/gve_adminq.c
index 1ced1e442e..2b25c7f390 100644
--- a/drivers/net/gve/base/gve_adminq.c
+++ b/drivers/net/gve/base/gve_adminq.c
@@ -263,6 +263,8 @@ int gve_adminq_alloc(struct gve_priv *priv)
priv->adminq_get_ptype_map_cnt = 0;
priv->adminq_cfg_flow_rule_cnt = 0;
+ priv->adminq_report_nic_timestamp_cnt = 0;
+
pthread_mutexattr_init(&mutexattr);
pthread_mutexattr_setpshared(&mutexattr, PTHREAD_PROCESS_SHARED);
pthread_mutex_init(&priv->adminq_lock, &mutexattr);
@@ -522,6 +524,10 @@ static int gve_adminq_issue_cmd(struct gve_priv *priv,
case GVE_ADMINQ_CONFIGURE_FLOW_RULE:
priv->adminq_cfg_flow_rule_cnt++;
break;
+ case GVE_ADMINQ_REPORT_NIC_TIMESTAMP:
+ priv->adminq_report_nic_timestamp_cnt++;
+ break;
+
default:
PMD_DRV_LOG(ERR, "unknown AQ command opcode %d", opcode);
}
@@ -636,6 +642,20 @@ int gve_adminq_reset_flow_rules(struct gve_priv *priv)
return gve_adminq_configure_flow_rule(priv, &flow_rule_cmd);
}
+int gve_adminq_report_nic_timestamp(struct gve_priv *priv, dma_addr_t nic_ts_report_addr)
+{
+ union gve_adminq_command cmd;
+
+ memset(&cmd, 0, sizeof(cmd));
+ cmd.opcode = cpu_to_be32(GVE_ADMINQ_REPORT_NIC_TIMESTAMP);
+ cmd.report_nic_timestamp = (struct gve_adminq_report_nic_timestamp) {
+ .nic_ts_report_len = cpu_to_be64(sizeof(struct gve_nic_ts_report)),
+ .nic_timestamp_addr = cpu_to_be64(nic_ts_report_addr),
+ };
+
+ return gve_adminq_execute_cmd(priv, &cmd);
+}
+
/* The device specifies that the management vector can either be the first irq
* or the last irq. ntfy_blk_msix_base_idx indicates the first irq assigned to
* the ntfy blks. It if is 0 then the management vector is last, if it is 1 then
diff --git a/drivers/net/gve/base/gve_adminq.h b/drivers/net/gve/base/gve_adminq.h
index eaee5649f2..954be39fbf 100644
--- a/drivers/net/gve/base/gve_adminq.h
+++ b/drivers/net/gve/base/gve_adminq.h
@@ -26,6 +26,7 @@ enum gve_adminq_opcodes {
GVE_ADMINQ_REPORT_LINK_SPEED = 0xD,
GVE_ADMINQ_GET_PTYPE_MAP = 0xE,
GVE_ADMINQ_VERIFY_DRIVER_COMPATIBILITY = 0xF,
+ GVE_ADMINQ_REPORT_NIC_TIMESTAMP = 0x11,
/* For commands that are larger than 56 bytes */
GVE_ADMINQ_EXTENDED_COMMAND = 0xFF,
};
@@ -373,6 +374,23 @@ struct gve_stats_report {
GVE_CHECK_STRUCT_LEN(8, gve_stats_report);
+struct gve_adminq_report_nic_timestamp {
+ __be64 nic_ts_report_len;
+ __be64 nic_timestamp_addr;
+};
+
+GVE_CHECK_STRUCT_LEN(16, gve_adminq_report_nic_timestamp);
+
+struct gve_nic_ts_report {
+ __be64 nic_timestamp; /* NIC clock in nanoseconds */
+ __be64 pre_cycles; /* System cycle counter before NIC clock read */
+ __be64 post_cycles; /* System cycle counter after NIC clock read */
+ __be64 reserved3;
+ __be64 reserved4;
+};
+
+GVE_CHECK_STRUCT_LEN(40, gve_nic_ts_report);
+
/* Numbers of gve tx/rx stats in stats report. */
#define GVE_TX_STATS_REPORT_NUM 6
#define GVE_RX_STATS_REPORT_NUM 2
@@ -490,6 +508,7 @@ union gve_adminq_command {
struct gve_adminq_verify_driver_compatibility
verify_driver_compatibility;
struct gve_adminq_extended_command extended_command;
+ struct gve_adminq_report_nic_timestamp report_nic_timestamp;
};
};
u8 reserved[64];
@@ -537,5 +556,6 @@ int gve_adminq_add_flow_rule(struct gve_priv *priv,
struct gve_flow_rule_params *rule, u32 loc);
int gve_adminq_del_flow_rule(struct gve_priv *priv, u32 loc);
int gve_adminq_reset_flow_rules(struct gve_priv *priv);
+int gve_adminq_report_nic_timestamp(struct gve_priv *priv, dma_addr_t nic_ts_report_addr);
#endif /* _GVE_ADMINQ_H */
diff --git a/drivers/net/gve/gve_ethdev.h b/drivers/net/gve/gve_ethdev.h
index b9b4688367..b67f82c263 100644
--- a/drivers/net/gve/gve_ethdev.h
+++ b/drivers/net/gve/gve_ethdev.h
@@ -328,6 +328,7 @@ struct gve_priv {
uint32_t adminq_get_ptype_map_cnt;
uint32_t adminq_verify_driver_compatibility_cnt;
uint32_t adminq_cfg_flow_rule_cnt;
+ uint32_t adminq_report_nic_timestamp_cnt;
volatile uint32_t state_flags;
/* Gvnic device link speed from hypervisor. */
--
2.54.0.1136.gdb2ca164c4-goog
^ permalink raw reply related
* [PATCH v4 2/6] net/gve: add device option support for HW timestamps
From: Mark Blasko @ 2026-06-13 4:22 UTC (permalink / raw)
To: stephen; +Cc: dev, joshwash, jtranoleary, blasko
In-Reply-To: <20260613042300.3760470-1-blasko@google.com>
Introduce the necessary definitions and functions for the device
option flag (GVE_DEV_OPT_ID_NIC_TIMESTAMP) to detect hardware
timestamping support in the gvnic device.
Signed-off-by: Mark Blasko <blasko@google.com>
Reviewed-by: Joshua Washington <joshwash@google.com>
Reviewed-by: Jasper Tran O'Leary <jtranoleary@google.com>
---
drivers/net/gve/base/gve_adminq.c | 41 ++++++++++++++++++++++++++-----
drivers/net/gve/base/gve_adminq.h | 9 +++++++
drivers/net/gve/gve_ethdev.h | 3 +++
3 files changed, 47 insertions(+), 6 deletions(-)
diff --git a/drivers/net/gve/base/gve_adminq.c b/drivers/net/gve/base/gve_adminq.c
index 743ab8e7ae..1ced1e442e 100644
--- a/drivers/net/gve/base/gve_adminq.c
+++ b/drivers/net/gve/base/gve_adminq.c
@@ -38,7 +38,8 @@ void gve_parse_device_option(struct gve_priv *priv,
struct gve_device_option_dqo_rda **dev_op_dqo_rda,
struct gve_device_option_flow_steering **dev_op_flow_steering,
struct gve_device_option_modify_ring **dev_op_modify_ring,
- struct gve_device_option_jumbo_frames **dev_op_jumbo_frames)
+ struct gve_device_option_jumbo_frames **dev_op_jumbo_frames,
+ struct gve_device_option_nic_timestamp **dev_op_nic_timestamp)
{
u32 req_feat_mask = be32_to_cpu(option->required_features_mask);
u16 option_length = be16_to_cpu(option->option_length);
@@ -168,6 +169,24 @@ void gve_parse_device_option(struct gve_priv *priv,
}
*dev_op_jumbo_frames = RTE_PTR_ADD(option, sizeof(*option));
break;
+ case GVE_DEV_OPT_ID_NIC_TIMESTAMP:
+ if (option_length < sizeof(**dev_op_nic_timestamp) ||
+ req_feat_mask != GVE_DEV_OPT_REQ_FEAT_MASK_NIC_TIMESTAMP) {
+ PMD_DRV_LOG(WARNING, GVE_DEVICE_OPTION_ERROR_FMT,
+ "Nic Timestamp",
+ (int)sizeof(**dev_op_nic_timestamp),
+ GVE_DEV_OPT_REQ_FEAT_MASK_NIC_TIMESTAMP,
+ option_length, req_feat_mask);
+ break;
+ }
+
+ if (option_length > sizeof(**dev_op_nic_timestamp)) {
+ PMD_DRV_LOG(WARNING,
+ GVE_DEVICE_OPTION_TOO_BIG_FMT,
+ "Nic Timestamp");
+ }
+ *dev_op_nic_timestamp = RTE_PTR_ADD(option, sizeof(*option));
+ break;
default:
/* If we don't recognize the option just continue
* without doing anything.
@@ -186,7 +205,8 @@ gve_process_device_options(struct gve_priv *priv,
struct gve_device_option_dqo_rda **dev_op_dqo_rda,
struct gve_device_option_flow_steering **dev_op_flow_steering,
struct gve_device_option_modify_ring **dev_op_modify_ring,
- struct gve_device_option_jumbo_frames **dev_op_jumbo_frames)
+ struct gve_device_option_jumbo_frames **dev_op_jumbo_frames,
+ struct gve_device_option_nic_timestamp **dev_op_nic_timestamp)
{
const int num_options = be16_to_cpu(descriptor->num_device_options);
struct gve_device_option *dev_opt;
@@ -207,7 +227,8 @@ gve_process_device_options(struct gve_priv *priv,
gve_parse_device_option(priv, dev_opt,
dev_op_gqi_rda, dev_op_gqi_qpl,
dev_op_dqo_rda, dev_op_flow_steering,
- dev_op_modify_ring, dev_op_jumbo_frames);
+ dev_op_modify_ring, dev_op_jumbo_frames,
+ dev_op_nic_timestamp);
dev_opt = next_opt;
}
@@ -920,7 +941,8 @@ static void gve_enable_supported_features(struct gve_priv *priv,
u32 supported_features_mask,
const struct gve_device_option_flow_steering *dev_op_flow_steering,
const struct gve_device_option_modify_ring *dev_op_modify_ring,
- const struct gve_device_option_jumbo_frames *dev_op_jumbo_frames)
+ const struct gve_device_option_jumbo_frames *dev_op_jumbo_frames,
+ const struct gve_device_option_nic_timestamp *dev_op_nic_timestamp)
{
if (dev_op_flow_steering &&
(supported_features_mask & GVE_SUP_FLOW_STEERING_MASK) &&
@@ -947,6 +969,11 @@ static void gve_enable_supported_features(struct gve_priv *priv,
PMD_DRV_LOG(INFO, "JUMBO FRAMES device option enabled.");
priv->max_mtu = be16_to_cpu(dev_op_jumbo_frames->max_mtu);
}
+ if (dev_op_nic_timestamp &&
+ (supported_features_mask & GVE_SUP_NIC_TIMESTAMP_MASK)) {
+ PMD_DRV_LOG(INFO, "NIC TIMESTAMP device option enabled.");
+ priv->nic_timestamp_supported = true;
+ }
}
int gve_adminq_describe_device(struct gve_priv *priv)
@@ -954,6 +981,7 @@ int gve_adminq_describe_device(struct gve_priv *priv)
struct gve_device_option_jumbo_frames *dev_op_jumbo_frames = NULL;
struct gve_device_option_modify_ring *dev_op_modify_ring = NULL;
struct gve_device_option_flow_steering *dev_op_flow_steering = NULL;
+ struct gve_device_option_nic_timestamp *dev_op_nic_timestamp = NULL;
struct gve_device_option_gqi_rda *dev_op_gqi_rda = NULL;
struct gve_device_option_gqi_qpl *dev_op_gqi_qpl = NULL;
struct gve_device_option_dqo_rda *dev_op_dqo_rda = NULL;
@@ -983,7 +1011,8 @@ int gve_adminq_describe_device(struct gve_priv *priv)
&dev_op_gqi_qpl, &dev_op_dqo_rda,
&dev_op_flow_steering,
&dev_op_modify_ring,
- &dev_op_jumbo_frames);
+ &dev_op_jumbo_frames,
+ &dev_op_nic_timestamp);
if (err)
goto free_device_descriptor;
@@ -1038,7 +1067,7 @@ int gve_adminq_describe_device(struct gve_priv *priv)
gve_enable_supported_features(priv, supported_features_mask,
dev_op_flow_steering, dev_op_modify_ring,
- dev_op_jumbo_frames);
+ dev_op_jumbo_frames, dev_op_nic_timestamp);
free_device_descriptor:
gve_free_dma_mem(&descriptor_dma_mem);
diff --git a/drivers/net/gve/base/gve_adminq.h b/drivers/net/gve/base/gve_adminq.h
index d8e5e6a352..eaee5649f2 100644
--- a/drivers/net/gve/base/gve_adminq.h
+++ b/drivers/net/gve/base/gve_adminq.h
@@ -153,6 +153,12 @@ struct gve_device_option_jumbo_frames {
GVE_CHECK_STRUCT_LEN(8, gve_device_option_jumbo_frames);
+struct gve_device_option_nic_timestamp {
+ __be32 supported_features_mask;
+};
+
+GVE_CHECK_STRUCT_LEN(4, gve_device_option_nic_timestamp);
+
/* Terminology:
*
* RDA - Raw DMA Addressing - Buffers associated with SKBs are directly DMA
@@ -169,6 +175,7 @@ enum gve_dev_opt_id {
GVE_DEV_OPT_ID_MODIFY_RING = 0x6,
GVE_DEV_OPT_ID_JUMBO_FRAMES = 0x8,
GVE_DEV_OPT_ID_FLOW_STEERING = 0xb,
+ GVE_DEV_OPT_ID_NIC_TIMESTAMP = 0xd,
};
enum gve_dev_opt_req_feat_mask {
@@ -179,12 +186,14 @@ enum gve_dev_opt_req_feat_mask {
GVE_DEV_OPT_REQ_FEAT_MASK_FLOW_STEERING = 0x0,
GVE_DEV_OPT_REQ_FEAT_MASK_MODIFY_RING = 0x0,
GVE_DEV_OPT_REQ_FEAT_MASK_JUMBO_FRAMES = 0x0,
+ GVE_DEV_OPT_REQ_FEAT_MASK_NIC_TIMESTAMP = 0x0,
};
enum gve_sup_feature_mask {
GVE_SUP_MODIFY_RING_MASK = 1 << 0,
GVE_SUP_JUMBO_FRAMES_MASK = 1 << 2,
GVE_SUP_FLOW_STEERING_MASK = 1 << 5,
+ GVE_SUP_NIC_TIMESTAMP_MASK = 1 << 8,
};
#define GVE_DEV_OPT_LEN_GQI_RAW_ADDRESSING 0x0
diff --git a/drivers/net/gve/gve_ethdev.h b/drivers/net/gve/gve_ethdev.h
index 524e48e723..b9b4688367 100644
--- a/drivers/net/gve/gve_ethdev.h
+++ b/drivers/net/gve/gve_ethdev.h
@@ -355,6 +355,9 @@ struct gve_priv {
void *avail_flow_rule_bmp_mem; /* Backing memory for the bitmap */
pthread_mutex_t flow_rule_lock; /* Lock for bitmap and tailq access */
TAILQ_HEAD(, gve_flow) active_flows;
+
+ /* HW Timestamping Fields */
+ bool nic_timestamp_supported;
};
static inline bool
--
2.54.0.1136.gdb2ca164c4-goog
^ permalink raw reply related
* [PATCH v4 1/6] net/gve: add thread safety to admin queue
From: Mark Blasko @ 2026-06-13 4:22 UTC (permalink / raw)
To: stephen; +Cc: dev, joshwash, jtranoleary, blasko
In-Reply-To: <20260613042300.3760470-1-blasko@google.com>
Introduce a pthread_mutex to protect the admin queue operations.
Locking was added around gve_adminq_execute_cmd and the batch
queue creation/destruction functions.
Signed-off-by: Mark Blasko <blasko@google.com>
Reviewed-by: Joshua Washington <joshwash@google.com>
Reviewed-by: Jasper Tran O'Leary <jtranoleary@google.com>
---
v2:
- Dropped ROBUST mutex attribute.
---
.mailmap | 1 +
drivers/net/gve/base/gve_adminq.c | 67 +++++++++++++++++++++++++------
drivers/net/gve/gve_ethdev.h | 1 +
3 files changed, 56 insertions(+), 13 deletions(-)
diff --git a/.mailmap b/.mailmap
index e052b85213..1b10cfca35 100644
--- a/.mailmap
+++ b/.mailmap
@@ -1016,6 +1016,7 @@ Mario Carrillo <mario.alfredo.c.arevalo@intel.com>
Mário Kuka <kuka@cesnet.cz>
Mariusz Drost <mariuszx.drost@intel.com>
Mark Asselstine <mark.asselstine@windriver.com>
+Mark Blasko <blasko@google.com>
Mark Bloch <mbloch@nvidia.com> <markb@mellanox.com>
Mark Gillott <mgillott@vyatta.att-mail.com>
Mark Kavanagh <mark.b.kavanagh@intel.com>
diff --git a/drivers/net/gve/base/gve_adminq.c b/drivers/net/gve/base/gve_adminq.c
index 9c5316fb00..743ab8e7ae 100644
--- a/drivers/net/gve/base/gve_adminq.c
+++ b/drivers/net/gve/base/gve_adminq.c
@@ -216,6 +216,7 @@ gve_process_device_options(struct gve_priv *priv,
int gve_adminq_alloc(struct gve_priv *priv)
{
+ pthread_mutexattr_t mutexattr;
uint8_t pci_rev_id;
priv->adminq = gve_alloc_dma_mem(&priv->adminq_dma_mem, PAGE_SIZE);
@@ -241,6 +242,11 @@ int gve_adminq_alloc(struct gve_priv *priv)
priv->adminq_get_ptype_map_cnt = 0;
priv->adminq_cfg_flow_rule_cnt = 0;
+ pthread_mutexattr_init(&mutexattr);
+ pthread_mutexattr_setpshared(&mutexattr, PTHREAD_PROCESS_SHARED);
+ pthread_mutex_init(&priv->adminq_lock, &mutexattr);
+ pthread_mutexattr_destroy(&mutexattr);
+
/* Setup Admin queue with the device */
rte_pci_read_config(priv->pci_dev, &pci_rev_id, sizeof(pci_rev_id),
RTE_PCI_REVISION_ID);
@@ -304,6 +310,7 @@ void gve_adminq_free(struct gve_priv *priv)
return;
gve_adminq_release(priv);
gve_free_dma_mem(&priv->adminq_dma_mem);
+ pthread_mutex_destroy(&priv->adminq_lock);
gve_clear_admin_queue_ok(priv);
}
@@ -418,7 +425,10 @@ static int gve_adminq_issue_cmd(struct gve_priv *priv,
(tail & priv->adminq_mask)) {
int err;
- /* Flush existing commands to make room. */
+ /* Flush existing commands to make room.
+ * Note: This kicks the doorbell for all staged commands.
+ * Any failure here means we failed after attempting to kick.
+ */
err = gve_adminq_kick_and_wait(priv);
if (err)
return err;
@@ -509,17 +519,24 @@ static int gve_adminq_execute_cmd(struct gve_priv *priv,
u32 tail, head;
int err;
+ pthread_mutex_lock(&priv->adminq_lock);
tail = ioread32be(&priv->reg_bar0->adminq_event_counter);
head = priv->adminq_prod_cnt;
- if (tail != head)
+ if (tail != head) {
/* This is not a valid path */
- return -EINVAL;
+ err = -EINVAL;
+ goto unlock_and_return;
+ }
err = gve_adminq_issue_cmd(priv, cmd_orig);
if (err)
- return err;
+ goto unlock_and_return;
- return gve_adminq_kick_and_wait(priv);
+ err = gve_adminq_kick_and_wait(priv);
+
+unlock_and_return:
+ pthread_mutex_unlock(&priv->adminq_lock);
+ return err;
}
static int gve_adminq_execute_extended_cmd(struct gve_priv *priv, u32 opcode,
@@ -693,13 +710,19 @@ int gve_adminq_create_tx_queues(struct gve_priv *priv, u32 num_queues)
int err;
u32 i;
+ pthread_mutex_lock(&priv->adminq_lock);
+
for (i = 0; i < num_queues; i++) {
err = gve_adminq_create_tx_queue(priv, i);
if (err)
- return err;
+ goto unlock_and_return;
}
- return gve_adminq_kick_and_wait(priv);
+ err = gve_adminq_kick_and_wait(priv);
+
+unlock_and_return:
+ pthread_mutex_unlock(&priv->adminq_lock);
+ return err;
}
static int gve_adminq_create_rx_queue(struct gve_priv *priv, u32 queue_index)
@@ -747,13 +770,19 @@ int gve_adminq_create_rx_queues(struct gve_priv *priv, u32 num_queues)
int err;
u32 i;
+ pthread_mutex_lock(&priv->adminq_lock);
+
for (i = 0; i < num_queues; i++) {
err = gve_adminq_create_rx_queue(priv, i);
if (err)
- return err;
+ goto unlock_and_return;
}
- return gve_adminq_kick_and_wait(priv);
+ err = gve_adminq_kick_and_wait(priv);
+
+unlock_and_return:
+ pthread_mutex_unlock(&priv->adminq_lock);
+ return err;
}
static int gve_adminq_destroy_tx_queue(struct gve_priv *priv, u32 queue_index)
@@ -779,13 +808,19 @@ int gve_adminq_destroy_tx_queues(struct gve_priv *priv, u32 num_queues)
int err;
u32 i;
+ pthread_mutex_lock(&priv->adminq_lock);
+
for (i = 0; i < num_queues; i++) {
err = gve_adminq_destroy_tx_queue(priv, i);
if (err)
- return err;
+ goto unlock_and_return;
}
- return gve_adminq_kick_and_wait(priv);
+ err = gve_adminq_kick_and_wait(priv);
+
+unlock_and_return:
+ pthread_mutex_unlock(&priv->adminq_lock);
+ return err;
}
static int gve_adminq_destroy_rx_queue(struct gve_priv *priv, u32 queue_index)
@@ -811,13 +846,19 @@ int gve_adminq_destroy_rx_queues(struct gve_priv *priv, u32 num_queues)
int err;
u32 i;
+ pthread_mutex_lock(&priv->adminq_lock);
+
for (i = 0; i < num_queues; i++) {
err = gve_adminq_destroy_rx_queue(priv, i);
if (err)
- return err;
+ goto unlock_and_return;
}
- return gve_adminq_kick_and_wait(priv);
+ err = gve_adminq_kick_and_wait(priv);
+
+unlock_and_return:
+ pthread_mutex_unlock(&priv->adminq_lock);
+ return err;
}
static int gve_set_desc_cnt(struct gve_priv *priv,
diff --git a/drivers/net/gve/gve_ethdev.h b/drivers/net/gve/gve_ethdev.h
index 0577f03974..524e48e723 100644
--- a/drivers/net/gve/gve_ethdev.h
+++ b/drivers/net/gve/gve_ethdev.h
@@ -339,6 +339,7 @@ struct gve_priv {
struct gve_tx_queue **txqs;
struct gve_rx_queue **rxqs;
+ pthread_mutex_t adminq_lock; /* Protects AdminQ command execution */
uint32_t stats_report_len;
const struct rte_memzone *stats_report_mem;
uint16_t stats_start_idx; /* start index of array of stats written by NIC */
--
2.54.0.1136.gdb2ca164c4-goog
^ permalink raw reply related
* [PATCH v4 0/6] net/gve: add hardware timestamping support
From: Mark Blasko @ 2026-06-13 4:22 UTC (permalink / raw)
To: stephen; +Cc: dev, joshwash, jtranoleary, blasko
In-Reply-To: <20260605213022.2770893-1-blasko@google.com>
This patch series introduces support for GVE hardware timestamping
on DQO queues. To support concurrent access, a mutex lock is introduced
to protect admin queue operations. A mechanism is then added to
periodically synchronize the NIC clock via a dedicated control thread,
and support is introduced for the read_clock ethdev operation.
Finally, the RX datapath is updated to reconstruct full 64-bit
timestamps from the 32-bit values in DQO descriptors.
---
v4:
- Patch 4: Offload periodic NIC clock reads to a dedicated control thread.
- Patch 5: Correct mutex initialization and teardown ordering.
- Patch 6: Update GVE documentation to clarify queue format requirements.
v3:
- Patch 5:
- Add mutex lock to protect shared NIC timestamp memzone access.
- Fix missing read_clock assignment to DQO queue ops table
(accidental omission in v2).
v2:
- Patch 1: Dropped ROBUST mutex attribute.
- Patch 3: Added adminq timestamp counter reset to gve_adminq_alloc.
- Patch 4:
- Removed redundant void* casts.
- Handled alarm reschedule failures by marking timestamp stale.
- Added transient error logging on memzone allocation failure.
- Patch 5: Scoped read_clock ethdev operation strictly to DQO queues.
- Patch 6:
- Scoped timestamp offload capability advertisement strictly to
DQO queues.
- Predicated capability advertisement directly on memzone
allocation.
- Initialized mbuf_timestamp_offset to -1.
- Added blank line separating release notes.
---
Mark Blasko (6):
net/gve: add thread safety to admin queue
net/gve: add device option support for HW timestamps
net/gve: add AdminQ command for NIC timestamps
net/gve: add periodic NIC clock synchronization
net/gve: support read clock ethdev op
net/gve: reconstruct HW timestamps from DQO
.mailmap | 1 +
doc/guides/nics/features/gve.ini | 1 +
doc/guides/nics/gve.rst | 20 +++
doc/guides/rel_notes/release_26_07.rst | 4 +
drivers/net/gve/base/gve_adminq.c | 128 +++++++++++++---
drivers/net/gve/base/gve_adminq.h | 29 ++++
drivers/net/gve/base/gve_desc_dqo.h | 8 +-
drivers/net/gve/gve_ethdev.c | 194 +++++++++++++++++++++++--
drivers/net/gve/gve_ethdev.h | 43 ++++++
drivers/net/gve/gve_rx_dqo.c | 26 ++++
10 files changed, 424 insertions(+), 30 deletions(-)
--
2.54.0.1136.gdb2ca164c4-goog
^ permalink raw reply
* [PATCH v2] tools: AI review handle empty Error sections
From: Matthew Gee @ 2026-06-12 19:08 UTC (permalink / raw)
To: dev; +Cc: stephen, aconole, lylavoie, Matthew Gee
In-Reply-To: <20260612190225.1016275-1-mgee@iol.unh.edu>
Previous review-patch.py would detect and report and error or warning
based off of the occurrence of the headers of the error and warning
sections. This led to consistent false positives as often AI reviewers
will create the header but put "none" or similar filler text within
the following body.
This patch updates the code in order to check if the AI review has a
body with error or warnings to fix and not just filler text. This is
done by querying multiple lines at once and adjusting the regex to
filter out headers followed only by filler text or formatting in the
review.
These changes were tested against 10+ AI review outputs with several
variations in formatting and filler text in order to catch a good
variety of cases to make sure code reviews with actual errors or
warnings are caught and not missed.
Signed-off-by: Matthew Gee <mgee@iol.unh.edu>
---
devtools/ai/review-patch.py | 35 +++++++++++++++++++++++++++--------
1 file changed, 27 insertions(+), 8 deletions(-)
diff --git a/devtools/ai/review-patch.py b/devtools/ai/review-patch.py
index 52601ac156..eb608566d2 100755
--- a/devtools/ai/review-patch.py
+++ b/devtools/ai/review-patch.py
@@ -19,6 +19,7 @@
from email.message import EmailMessage
from pathlib import Path
from typing import Any, Iterator
+from itertools import tee, islice, chain
from _common import (
PROVIDERS,
@@ -147,19 +148,37 @@ def classify_review(review_text: str, output_format: str) -> int:
pass # Fall through to text scanning
if not has_errors and not has_warnings:
- # Scan review text for severity indicators.
- # Match section headers and inline markers across text/markdown/html.
- for line in review_text.splitlines():
- stripped = line.strip().lower()
+ # Matches against error or warning section headers
+ rgx_should_match: str = r"#+\s(\*+)?{err_or_warn}"
+ # Matches against headers followed only by filler text, formatting, or no additional text
+ rgx_should_not_match: str = r"#+\s(\*+)?{err_or_warn}(s?)(\*+)?(none(.)?$| \(must fix\)$|$)"
+
+ curr_lines: iter[str]
+ next_lines: iter[str | None]
+ next_next_lines: iter[str | None]
+ curr_lines, next_lines, next_next_lines = tee(review_text.splitlines(), 3)
+ next_lines = chain(islice(next_lines, 1, None), [None])
+ next_next_lines = chain(islice(next_next_lines, 2, None), [None, None])
+
+ curr_lines: str
+ next_lines: str | None
+ next_next_lines: str | None
+ for curr_line, next_line, next_next_line in zip(curr_lines, next_lines, next_next_lines):
+ stripped: str = curr_line.strip().lower() + str(
+ next_line or '').strip().lower() + str(next_next_line or '').strip().lower()
# Skip quoted patch context lines
if stripped.startswith(">") or stripped.startswith("diff --git"):
continue
- if re.match(r"^(#{1,3}\s+)?(\*{0,2})error", stripped) or re.match(
- r"^<h[1-3]>\s*error", stripped
+
+ elif re.match(rgx_should_match.format(err_or_warn='error'),
+ stripped) and not re.match(
+ rgx_should_not_match.format(err_or_warn='error'), stripped
):
has_errors = True
- elif re.match(r"^(#{1,3}\s+)?(\*{0,2})warning", stripped) or re.match(
- r"^<h[1-3]>\s*warning", stripped
+
+ elif re.match(rgx_should_match.format(err_or_warn='warning'),
+ stripped) and not re.match(
+ rgx_should_not_match.format(err_or_warn='warning'), stripped
):
has_warnings = True
--
2.54.0
^ permalink raw reply related
* [PATCH v1] tools: AI review handle empty Error sections
From: Matthew Gee @ 2026-06-12 19:02 UTC (permalink / raw)
To: dev; +Cc: stephen, aconole, lylavoie, Matthew Gee
Previous review-patch.py would detect and report and error or warning
based off of the occurrence of the headers of the error and warning
sections. This led to consistent false positives as often AI reviewers
will create the header but put "none" or similar filler text within
the following body.
This patch updates the code in order to check if the AI review has a
body with error or warnings to fix and not just filler text. This is
done by querying multiple lines at once and adjusting the regex to
filter out headers followed only by filler text or formatting in the
review.
These changes were tested against 10+ AI review outputs with several
variations in formatting and filler text in order to catch a good
variety of cases to make sure code reviews with actual errors or
warnings are caught and not missed.
Signed-off-by: Matthew Gee <mgee@iol.unh.edu>
---
devtools/ai/review-patch.py | 35 +++++++++++++++++++++++++++--------
1 file changed, 27 insertions(+), 8 deletions(-)
diff --git a/devtools/ai/review-patch.py b/devtools/ai/review-patch.py
index 52601ac156..b009368638 100755
--- a/devtools/ai/review-patch.py
+++ b/devtools/ai/review-patch.py
@@ -19,6 +19,7 @@
from email.message import EmailMessage
from pathlib import Path
from typing import Any, Iterator
+from itertools import tee, islice, chain
from _common import (
PROVIDERS,
@@ -147,19 +148,37 @@ def classify_review(review_text: str, output_format: str) -> int:
pass # Fall through to text scanning
if not has_errors and not has_warnings:
- # Scan review text for severity indicators.
- # Match section headers and inline markers across text/markdown/html.
- for line in review_text.splitlines():
- stripped = line.strip().lower()
+ # Matches against error or warning section headers
+ rgx_should_match: str = r"#+\s(\*+)?{err_or_warn}"
+ # Matches against headers followed only by filler text, formatting, or no additional text
+ rgx_should_not_match: str = r"#+\s(\*+)?{err_or_warn}(s?)(\*+)?(none(.)?$| \(must fix\)$|$)"
+
+ curr_lines: iter[str]
+ next_lines: iter[str | None]
+ next_next_lines: iter[str | None]
+ curr_lines, next_lines, next_next_lines = tee(review_text.splitlines(), 3)
+ next_lines = chain(islice(next_lines, 1, None), [None])
+ next_next_lines = chain(islice(next_next_lines, 2, None), [None, None])
+
+ curr_lines: str
+ next_lines: str | None
+ next_next_lines: str | None
+ for curr_line, next_line, next_next_line in zip(curr_lines, next_lines, next_next_lines):
+ stripped: str = curr_line.strip().lower() + str(
+ next_line or '').strip().lower() + str(next_next_line or '').strip().lower()
# Skip quoted patch context lines
if stripped.startswith(">") or stripped.startswith("diff --git"):
continue
- if re.match(r"^(#{1,3}\s+)?(\*{0,2})error", stripped) or re.match(
- r"^<h[1-3]>\s*error", stripped
+
+ elif re.match(rgx_should_match.format(err_or_warn='error'),
+ stripped) and not re.match(
+ rgx_should_not_match.format(err_or_warn='error'), stripped
):
has_errors = True
- elif re.match(r"^(#{1,3}\s+)?(\*{0,2})warning", stripped) or re.match(
- r"^<h[1-3]>\s*warning", stripped
+
+ elif re.match(rgx_should_match.format(err_or_warn='warning'),
+ stripped) and not re.match(
+ rgx_should_not_match.format(err_or_warn='warning'), stripped
):
has_warnings = True
--
2.54.0
^ permalink raw reply related
* Re: [PATCH v2] app/testpmd: add padding mode to txonly engine
From: Stephen Hemminger @ 2026-06-12 16:13 UTC (permalink / raw)
To: Xingui Yang
Cc: dev, david.marchand, aman.deep.singh, fengchengwen, yangshuaisong,
lihuisong, liuyonglong, kangfenglong
In-Reply-To: <20260612091217.2899755-1-yangxingui@huawei.com>
On Fri, 12 Jun 2026 17:12:17 +0800
Xingui Yang <yangxingui@huawei.com> wrote:
> Add a new padding mode to the txonly forwarding engine, which allows
> sending packets with configurable small sizes without standard L2/L3
> headers. This is useful for testing NIC padding logic.
>
> When padding mode is enabled via --tx-pkt-pad-mode flag:
> - l2_len and l3_len are set to 0 instead of standard header lengths
> - Packet data is filled with a static pattern instead of
> Ethernet/IP/UDP headers
> - Minimum packet length validation is bypassed to allow small
> packet sizes (e.g., set txpkts 14)
>
> Signed-off-by: Xingui Yang <yangxingui@huawei.com>
> Signed-off-by: Huisong Li <lihuisong@huawei.com>
> ---
> v2: Fix compilation exception of unterminated-string-initialization
> ---
What about something like this (*not tested*) patch.
From 84ff35849f9881a93eed65ccd43a5cd1197cecb8 Mon Sep 17 00:00:00 2001
From: Stephen Hemminger <stephen@networkplumber.org>
Date: Fri, 12 Jun 2026 09:10:17 -0700
Subject: [PATCH] testpnd: allow configuring runt frames
Allow setting transmit size to be a small value which has
ethernet header but no IP or UDP header.
---
app/test-pmd/config.c | 12 +++----
app/test-pmd/txonly.c | 35 +++++++++++++++++++--
doc/guides/testpmd_app_ug/testpmd_funcs.rst | 13 ++++++++
3 files changed, 51 insertions(+), 9 deletions(-)
diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c
index 9d457ca88e..46ff678b9f 100644
--- a/app/test-pmd/config.c
+++ b/app/test-pmd/config.c
@@ -6327,9 +6327,6 @@ set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs)
/*
* Check that each segment length is greater or equal than
* the mbuf data size.
- * Check also that the total packet length is greater or equal than the
- * size of an empty UDP/IP packet (sizeof(struct rte_ether_hdr) +
- * 20 + 8).
*/
tx_pkt_len = 0;
for (i = 0; i < nb_segs; i++) {
@@ -6341,10 +6338,11 @@ set_tx_pkt_segments(unsigned int *seg_lengths, unsigned int nb_segs)
}
tx_pkt_len = (uint16_t)(tx_pkt_len + seg_lengths[i]);
}
- if (tx_pkt_len < (sizeof(struct rte_ether_hdr) + 20 + 8)) {
- fprintf(stderr, "total packet length=%u < %d - give up\n",
- (unsigned) tx_pkt_len,
- (int)(sizeof(struct rte_ether_hdr) + 20 + 8));
+
+ /* Allow runt packets this is a test tool. */
+ if (tx_pkt_len < (sizeof(struct rte_ether_hdr))) {
+ fprintf(stderr, "total packet length=%u < %zu - give up\n",
+ tx_pkt_len, sizeof(struct rte_ether_hdr));
return;
}
diff --git a/app/test-pmd/txonly.c b/app/test-pmd/txonly.c
index 64893fa205..2e0abe361e 100644
--- a/app/test-pmd/txonly.c
+++ b/app/test-pmd/txonly.c
@@ -76,6 +76,12 @@ copy_buf_to_pkt_segs(void* buf, unsigned len, struct rte_mbuf *pkt,
while (offset >= seg->data_len) {
offset -= seg->data_len;
seg = seg->next;
+ /*
+ * The packet may be shorter than the header stack when
+ * generating runt frames; stop once it runs out of segments.
+ */
+ if (seg == NULL)
+ return;
}
copy_len = seg->data_len - offset;
seg_buf = rte_pktmbuf_mtod_offset(seg, char *, offset);
@@ -84,6 +90,8 @@ copy_buf_to_pkt_segs(void* buf, unsigned len, struct rte_mbuf *pkt,
len -= copy_len;
buf = ((char*) buf + copy_len);
seg = seg->next;
+ if (seg == NULL)
+ return;
seg_buf = rte_pktmbuf_mtod(seg, char *);
copy_len = seg->data_len;
}
@@ -193,7 +201,6 @@ pkt_burst_prepare(struct rte_mbuf *pkt, struct rte_mempool *mbp,
pkt->vlan_tci = vlan_tci;
pkt->vlan_tci_outer = vlan_tci_outer;
pkt->l2_len = sizeof(struct rte_ether_hdr);
- pkt->l3_len = sizeof(struct rte_ipv4_hdr);
pkt_len = pkt->data_len;
pkt_seg = pkt;
@@ -204,6 +211,24 @@ pkt_burst_prepare(struct rte_mbuf *pkt, struct rte_mempool *mbp,
pkt_len += pkt_seg->data_len;
}
pkt_seg->next = NULL; /* Last segment of packet. */
+
+ /*
+ * A runt frame may be too short to carry a full IPv4/UDP header.
+ * Clamp l3_len and drop any checksum offload whose header is not
+ * fully present, so the PMD is never asked to checksum bytes that
+ * are not in the frame. pkt_len is at least sizeof(rte_ether_hdr),
+ * so the subtraction below cannot underflow.
+ */
+ pkt->l3_len = RTE_MIN(sizeof(struct rte_ipv4_hdr),
+ pkt_len - sizeof(struct rte_ether_hdr));
+ if (pkt_len < sizeof(struct rte_ether_hdr) +
+ sizeof(struct rte_ipv4_hdr))
+ pkt->ol_flags &= ~(RTE_MBUF_F_TX_IP_CKSUM |
+ RTE_MBUF_F_TX_L4_MASK);
+ else if (pkt_len < sizeof(struct rte_ether_hdr) +
+ sizeof(struct rte_ipv4_hdr) +
+ sizeof(struct rte_udp_hdr))
+ pkt->ol_flags &= ~RTE_MBUF_F_TX_L4_MASK;
/*
* Copy headers in first packet segment(s).
*/
@@ -405,7 +430,13 @@ tx_only_begin(portid_t pi)
pkt_hdr_len = (uint16_t)(sizeof(struct rte_ether_hdr) +
sizeof(struct rte_ipv4_hdr) +
sizeof(struct rte_udp_hdr));
- pkt_data_len = tx_pkt_length - pkt_hdr_len;
+ /*
+ * tx_pkt_length may be smaller than the full header stack when
+ * generating runt frames; clamp the payload length to zero in that
+ * case so the IP/UDP length fields stay sane.
+ */
+ pkt_data_len = tx_pkt_length > pkt_hdr_len ?
+ tx_pkt_length - pkt_hdr_len : 0;
if ((tx_pkt_split == TX_PKT_SPLIT_RND || txonly_multi_flow) &&
tx_pkt_seg_lengths[0] < pkt_hdr_len) {
diff --git a/doc/guides/testpmd_app_ug/testpmd_funcs.rst b/doc/guides/testpmd_app_ug/testpmd_funcs.rst
index f0f2b0758b..d2e5b63586 100644
--- a/doc/guides/testpmd_app_ug/testpmd_funcs.rst
+++ b/doc/guides/testpmd_app_ug/testpmd_funcs.rst
@@ -876,6 +876,19 @@ Set the length of each segment of the TX-ONLY packets or length of packet for FL
Where x[,y]* represents a CSV list of values, without white space.
+The total packet length may be set as small as the Ethernet header
+(``sizeof(struct rte_ether_hdr)``), which is below the size of an empty
+UDP/IPv4 packet. This generates runt frames with a truncated (or absent)
+IPv4/UDP header, which is useful for testing how a driver handles
+undersized frames. Any header bytes that do not fit in the requested
+length are simply not emitted. Checksum offloads are automatically
+dropped for a frame too short to contain the corresponding header.
+
+Note that random split (``set txsplit rand``) and multi-flow
+(``set txonly-flows``) still require the first segment to hold the full
+Ethernet/IPv4/UDP header stack, so they cannot be combined with runt
+lengths.
+
set txtimes
~~~~~~~~~~~
--
2.53.0
^ permalink raw reply related
* Re: [PATCH v5 00/11] bpf: introduce extensible load API
From: Stephen Hemminger @ 2026-06-12 16:06 UTC (permalink / raw)
To: Marat Khalili; +Cc: dev
In-Reply-To: <20260612084219.38399-1-marat.khalili@huawei.com>
On Fri, 12 Jun 2026 09:42:07 +0100
Marat Khalili <marat.khalili@huawei.com> wrote:
> This patchset introduces an extensible load API for the BPF library in
> DPDK, addressing current limitations regarding ABI stability and feature
> constraints.
>
> Currently, `rte_bpf_load` relies on a fixed `struct rte_bpf_prm`, which
> makes it difficult to add new loading options or parameters without
> breaking the ABI.
>
> To resolve these issues, this series introduces `rte_bpf_load_ex` taking
> `struct rte_bpf_prm_ex`. The new parameter structure includes a `sz`
> field for backward compatibility, allowing future extensions.
>
> Taking advantage of the new extensible API, this patchset also adds
> several new features:
> * Support for loading and executing BPF programs with up to 5 arguments.
> * Support for loading classic BPF (cBPF) directly.
> * Support for loading ELF files directly from memory buffers.
> * New API functions (`rte_bpf_eth_rx_install` and `rte_bpf_eth_tx_install`)
> to install an already loaded BPF program as a port callback, decoupling
> the loading phase from the installation phase.
>
>
> v5:
> * Fixed compilation between commits broken in v4 while addressing AI
> comments.
> * Rebased on fresh main, addressing conflicts in release notes.
>
> v4:
> * Restored missing NULL checks in wrapper functions `rte_bpf_load` and
> `rte_bpf_elf_load`.
> * Fixed the burst execution functions (`rte_bpf_exec_burst*`) to return
> `0` and set `rte_errno = EINVAL` on failure, preventing `-EINVAL`
> being reinterpreted as a large `uint32_t` value. Initialized `rc`
> properly in scalar execution wrappers for this case.
> * Swapped the Doxygen comments in `rte_bpf_ethdev.h` for RX and TX functions.
> * Added diagnostic dump on failure path in `test_bpf_filter`.
> * Fixed memory leak of the BPF handle in `bpf_rx_test` upon install failure.
> * Added tests for NULL parameter rejection, mismatched execution arguments,
> unsupported execution flags, and the libpcap-less `rte_bpf_convert` stub.
>
> v3:
> * Appended Acked-by tags to all individual commits to align with
> patchwork requirements.
>
> v2:
> * Fixed a potential segmentation fault in `exec_vm_burst_ex` by deferring
> the dereference of `ctx[i].arg` until it is confirmed that `nb_prog_arg > 0`.
> * Clarified documentation and code comments for `RTE_BPF_EXEC_FLAG_JIT`
> requirements and fast-path expectations.
>
>
> Marat Khalili (11):
> bpf: make logging prefixes more consistent
> bpf: introduce extensible load API
> bpf: support up to 5 arguments
> bpf: add cBPF origin to rte_bpf_load_ex
> bpf: support rte_bpf_prm_ex with port callbacks
> bpf: support loading ELF files from memory
> test/bpf: test loading cBPF directly
> test/bpf: test loading ELF file from memory
> doc: add release notes for new extensible BPF API
> doc: add load API to BPF programmer's guide
> test/bpf: add tests for error handling contracts
>
> app/test/test_bpf.c | 455 +++++++++++++++++--------
> doc/guides/prog_guide/bpf_lib.rst | 75 +++-
> doc/guides/rel_notes/release_26_07.rst | 20 ++
> lib/bpf/bpf.c | 32 +-
> lib/bpf/bpf_convert.c | 99 +++++-
> lib/bpf/bpf_exec.c | 134 +++++++-
> lib/bpf/bpf_impl.h | 53 ++-
> lib/bpf/bpf_jit_arm64.c | 18 +-
> lib/bpf/bpf_jit_x86.c | 10 +-
> lib/bpf/bpf_load.c | 213 ++++++++++--
> lib/bpf/bpf_load_elf.c | 189 ++++++----
> lib/bpf/bpf_pkt.c | 65 +++-
> lib/bpf/bpf_stub.c | 46 ---
> lib/bpf/bpf_validate.c | 94 +++--
> lib/bpf/meson.build | 15 +-
> lib/bpf/rte_bpf.h | 199 ++++++++++-
> lib/bpf/rte_bpf_ethdev.h | 54 +++
> 17 files changed, 1400 insertions(+), 371 deletions(-)
> delete mode 100644 lib/bpf/bpf_stub.c
>
Detailed AI review found some minor things:
02/11: comment grammar nits — "Any features that not known to
the application" -> "that are not known" (twice); "the size of
same struct" -> "the size of the same struct".
05/11: Doxygen for rte_bpf_eth_rx_install and
rte_bpf_eth_tx_install references rte_bpf_eth_unload, which does
not exist. Use rte_bpf_eth_rx_unload and rte_bpf_eth_tx_unload
respectively.
07/11: stray space before comma in the test_bpf_filter log
string ("for \"%s\" ,").
^ permalink raw reply
* Re: [PATCH] net/iavf: fix scalar Rx path zero-length segment
From: Bruce Richardson @ 2026-06-12 15:42 UTC (permalink / raw)
To: Ciara Loftus; +Cc: dev, stable, Declan Doherty
In-Reply-To: <20260612143531.2265914-1-ciara.loftus@intel.com>
On Fri, Jun 12, 2026 at 02:35:31PM +0000, Ciara Loftus wrote:
> When hardware CRC stripping is active, a frame whose on-wire size is an
> exact multiple of the Rx buffer size can cause the NIC to fill the final
> data descriptor and place the four CRC bytes into a separate trailing
> descriptor. After hardware stripping, that descriptor carries zero bytes
> of payload.
>
> The existing CRC cleanup code only handles a zero-length trailing segment
> when software CRC stripping is enabled. When hardware stripping is
> active, the zero-length mbuf is silently chained to the reassembled
> packet. Forwarding such a packet causes a zero-length Tx descriptor,
> triggering a Malicious Driver Detection event on the PF and resetting
> the VF.
>
> Fix by adding logic to detect a zero-length final segment when hardware
> CRC stripping is active, and freeing it.
>
> Fixes: a2b29a7733ef ("net/avf: enable basic Rx Tx")
> Fixes: b8b4c54ef9b0 ("net/iavf: support flexible Rx descriptor in normal path")
> Cc: stable@dpdk.org
>
> Signed-off-by: Declan Doherty <declan.doherty@intel.com>
> Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
> ---
> drivers/net/intel/iavf/iavf_rxtx.c | 16 ++++++++++++++++
> 1 file changed, 16 insertions(+)
>
> diff --git a/drivers/net/intel/iavf/iavf_rxtx.c b/drivers/net/intel/iavf/iavf_rxtx.c
> index a57af7faed..86ebb2618d 100644
> --- a/drivers/net/intel/iavf/iavf_rxtx.c
> +++ b/drivers/net/intel/iavf/iavf_rxtx.c
> @@ -1716,6 +1716,14 @@ iavf_recv_scattered_pkts_flex_rxd(void *rx_queue, struct rte_mbuf **rx_pkts,
> rxm->data_len = (uint16_t)(rx_packet_len -
> RTE_ETHER_CRC_LEN);
> }
> + } else if (unlikely(rx_packet_len == 0)) {
> + /*
> + * NIC split CRC bytes into a trailing segment which is
> + * now empty after hardware CRC stripping. Free it.
> + */
> + rte_pktmbuf_free_seg(rxm);
> + first_seg->nb_segs--;
> + last_seg->next = NULL;
> }
>
The vector paths also handle scattered packets (via reassembly). Do they
need a fix for this? What about the other drivers that work on the PF, such
as ice/i40e?
/Bruce
> first_seg->port = rxq->port_id;
> @@ -1884,6 +1892,14 @@ iavf_recv_scattered_pkts(void *rx_queue, struct rte_mbuf **rx_pkts,
> } else
> rxm->data_len = (uint16_t)(rx_packet_len -
> RTE_ETHER_CRC_LEN);
> + } else if (unlikely(rx_packet_len == 0)) {
> + /*
> + * NIC split CRC bytes into a trailing segment which is
> + * now empty after hardware CRC stripping. Free it.
> + */
> + rte_pktmbuf_free_seg(rxm);
> + first_seg->nb_segs--;
> + last_seg->next = NULL;
> }
>
> first_seg->port = rxq->port_id;
> --
> 2.43.0
>
^ permalink raw reply
* Re: [PATCH] app/testpmd: add padding mode to txonly engine
From: Stephen Hemminger @ 2026-06-12 15:20 UTC (permalink / raw)
To: Xingui Yang
Cc: dev, david.marchand, aman.deep.singh, fengchengwen, yangshuaisong,
lihuisong, liuyonglong, kangfenglong
In-Reply-To: <20260612073715.2739007-1-yangxingui@huawei.com>
On Fri, 12 Jun 2026 15:37:15 +0800
Xingui Yang <yangxingui@huawei.com> wrote:
> Add a new padding mode to the txonly forwarding engine, which allows
> sending packets with configurable small sizes without standard L2/L3
> headers. This is useful for testing NIC padding logic.
>
> When padding mode is enabled via --tx-pkt-pad-mode flag:
> - l2_len and l3_len are set to 0 instead of standard header lengths
> - Packet data is filled with a static pattern instead of
> Ethernet/IP/UDP headers
> - Minimum packet length validation is bypassed to allow small
> packet sizes (e.g., set txpkts 14)
>
> Signed-off-by: Xingui Yang <yangxingui@huawei.com>
> Signed-off-by: Huisong Li <lihuisong@huawei.com>
> ---
Why add yet another setting to already bloated testpmd command?
Instead I would suggest allowing user to specify any length from
14 up to UINT32_MAX. The code to format packet would need to
handle it there.
^ permalink raw reply
* Re: [PATCH 8/9] ethdev: keep fast-path ops valid after port stop
From: Thomas Monjalon @ 2026-06-12 15:00 UTC (permalink / raw)
To: Maxime Leroy
Cc: Morten Brørup, Hemant Agrawal, Sachin Saxena, dev, stable,
Andrew Rybchenko, Sunil Kumar Kori
In-Reply-To: <CAHHRULUGdnTmpnAX2aONEMj3bJmN11RG2cHGJ7h3-qhjoJLsFw@mail.gmail.com>
11/06/2026 20:39, Maxime Leroy:
> Le jeu. 11 juin 2026, 18:01, Morten Brørup <mb@smartsharesystems.com> a
> écrit :
>
> > > From: Maxime Leroy [mailto:maxime.leroys@gmail.com] On Behalf Of Maxime
> > > Leroy
> >
> > Good catch.
> > Acked-by: Morten Brørup <mb@smartsharesystems.com>
> >
> > Not related to the series, consider sending as separate patch.
> >
> Thanks for the review and Ack.
>
> Agreed, this is a generic ethdev fix. I kept it in this series because the
> NAPI user depends on it.
>
> The current Grout NAPI loop arms RX queue interrupts and then re-checks
> rte_eth_rx_queue_count() before blocking, to avoid sleeping when a packet
> arrived between the last empty poll and epoll_wait.
>
> With the current ethdev reset path, rx_burst is replaced by a dummy
> callback on stop/release, but rx_queue_count becomes NULL. So if the port
> is stopped concurrently, the NAPI worker dereferences a NULL function
> pointer and
> segfaults on that recheck.
>
> I can split it out if maintainers prefer, but then the dpaa2 NAPI series
> has a real dependency on the standalone ethdev fix.
You can keep this patch in the series AND
send it separately for a quick merge.
^ permalink raw reply
* Re: 回复：[PATCH] gpu/metax: add new driver for Metax GPU
From: Thomas Monjalon @ 2026-06-12 14:49 UTC (permalink / raw)
To: 许玲燕; +Cc: dev, eagostini, 王冬冬
In-Reply-To: <2a306d48-351c-473e-b049-43e10e53617b.lingyan.xu@metax-tech.com>
Would it be possible to have a direct download link
not requiring any registration? At least for the library?
Do you have an english version of the website?
12/06/2026 09:19, 许玲燕:
> Hi,
> Thank you for the follow-up. Please find the clarifications below regarding the kernel dependency and availability:
> 1. Kernel Dependency
> While it is not upstreamed into the mainline Linux kernel, it is actively maintained by Metax to interface with our hardware.
> 2. Availability and Download Link
> Yes, both the proprietary user-space libraries and the corresponding kernel modules are freely available for download. You can access them via the Metax Software Download Center:
> Download Link: https://sw-download.metax-tech.com/ <https://sw-download.metax-tech.com/ >
> How to obtain the files:
>
> *
> Register and log in to the portal.
>
> *
> Navigate to "SDK Development Kit".
>
> *
> Select your specific GPU Type.
>
> *
> Choose your target OS. We currently support Linux (aarch64 & x86_64) across mainstream distributions, including but not limited to:
> *
> Ubuntu (18.04 - 24.04)
>
> *
> RHEL / CentOS / RockyOS (8.x / 9.x)
>
> *
> Domestic distros: KylinV10/V11, OpenCloudOS, TencentOS, etc.
> Best regards,
> ------------------------------------------------------------------
> 发件人：Thomas Monjalon <thomas@monjalon.net>
> 发送时间：2026年6月11日(周四) 17:17
> 收件人："许玲燕"<lingyan.xu@metax-tech.com>
> 抄 送：dev<dev@dpdk.org>; eagostini<eagostini@nvidia.com>; "王冬冬"<dongdong.wang@metax-tech.com>
> 主 题：Re: [PATCH] gpu/metax: add new driver for Metax GPU
> 11/06/2026 09:10:
> > Both libmcruntime.so and the corresponding gdrapi libraries
> > are proprietary user-space libraries provided by Metax.
> > They are not upstreamed to the DPDK mainline repository.
> > However, please rest assured that the current patch interacts
> > with them via standard dlopen (dynamic loading) at runtime.
> > We do not link directly against their source code
> > or require them as hard build-time dependencies.
> > Therefore, this approach will not introduce any additional
> > compilation dependencies or licensing issues to the DPDK main tree.
> What about the kernel dependency?
> Are libraries and kernel module freely available for download?
> Can you provide a link?
>
^ permalink raw reply
* [PATCH] net/iavf: fix scalar Rx path zero-length segment
From: Ciara Loftus @ 2026-06-12 14:35 UTC (permalink / raw)
To: dev; +Cc: Ciara Loftus, stable, Declan Doherty
When hardware CRC stripping is active, a frame whose on-wire size is an
exact multiple of the Rx buffer size can cause the NIC to fill the final
data descriptor and place the four CRC bytes into a separate trailing
descriptor. After hardware stripping, that descriptor carries zero bytes
of payload.
The existing CRC cleanup code only handles a zero-length trailing segment
when software CRC stripping is enabled. When hardware stripping is
active, the zero-length mbuf is silently chained to the reassembled
packet. Forwarding such a packet causes a zero-length Tx descriptor,
triggering a Malicious Driver Detection event on the PF and resetting
the VF.
Fix by adding logic to detect a zero-length final segment when hardware
CRC stripping is active, and freeing it.
Fixes: a2b29a7733ef ("net/avf: enable basic Rx Tx")
Fixes: b8b4c54ef9b0 ("net/iavf: support flexible Rx descriptor in normal path")
Cc: stable@dpdk.org
Signed-off-by: Declan Doherty <declan.doherty@intel.com>
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
---
drivers/net/intel/iavf/iavf_rxtx.c | 16 ++++++++++++++++
1 file changed, 16 insertions(+)
diff --git a/drivers/net/intel/iavf/iavf_rxtx.c b/drivers/net/intel/iavf/iavf_rxtx.c
index a57af7faed..86ebb2618d 100644
--- a/drivers/net/intel/iavf/iavf_rxtx.c
+++ b/drivers/net/intel/iavf/iavf_rxtx.c
@@ -1716,6 +1716,14 @@ iavf_recv_scattered_pkts_flex_rxd(void *rx_queue, struct rte_mbuf **rx_pkts,
rxm->data_len = (uint16_t)(rx_packet_len -
RTE_ETHER_CRC_LEN);
}
+ } else if (unlikely(rx_packet_len == 0)) {
+ /*
+ * NIC split CRC bytes into a trailing segment which is
+ * now empty after hardware CRC stripping. Free it.
+ */
+ rte_pktmbuf_free_seg(rxm);
+ first_seg->nb_segs--;
+ last_seg->next = NULL;
}
first_seg->port = rxq->port_id;
@@ -1884,6 +1892,14 @@ iavf_recv_scattered_pkts(void *rx_queue, struct rte_mbuf **rx_pkts,
} else
rxm->data_len = (uint16_t)(rx_packet_len -
RTE_ETHER_CRC_LEN);
+ } else if (unlikely(rx_packet_len == 0)) {
+ /*
+ * NIC split CRC bytes into a trailing segment which is
+ * now empty after hardware CRC stripping. Free it.
+ */
+ rte_pktmbuf_free_seg(rxm);
+ first_seg->nb_segs--;
+ last_seg->next = NULL;
}
first_seg->port = rxq->port_id;
--
2.43.0
^ permalink raw reply related
* RE: [PATCH v4 02/11] bpf: introduce extensible load API
From: Marat Khalili @ 2026-06-12 10:48 UTC (permalink / raw)
To: Thomas Monjalon; +Cc: Konstantin Ananyev, Wathsala Vithanage, dev@dpdk.org
In-Reply-To: <n9qJkNlCQDazx_P9i-5XTQ@monjalon.net>
> Unfortunately, compilation is failing on this patch:
>
> lib/bpf/bpf_load.c:274:40: error: 'struct rte_bpf_jit' has no member named 'raw'
Sent v5 fixing the issue, apologies for sending without double-checking
compilability of individual commits on my side.
^ permalink raw reply
* [PATCH v3 25/25] doc: add BPF validate debug to programmer's guide
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Document the new gdb-like validation debugger API, outlining how it can
be used to set breakpoints and inspect register states during
validation.
Highlight its primary use case: writing robust tests for the eBPF
verifier using the harness in app/test/test_bpf_validate.c.
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
doc/guides/prog_guide/bpf_lib.rst | 31 +++++++++++++++++++++++++++++++
1 file changed, 31 insertions(+)
diff --git a/doc/guides/prog_guide/bpf_lib.rst b/doc/guides/prog_guide/bpf_lib.rst
index df3782508829..08ea1876875a 100644
--- a/doc/guides/prog_guide/bpf_lib.rst
+++ b/doc/guides/prog_guide/bpf_lib.rst
@@ -118,6 +118,37 @@ For example, ``(BPF_IND | BPF_W | BPF_LD)`` means:
and ``R1-R5`` were scratched.
+Validation Debugger
+-------------------
+
+The DPDK BPF library includes a validation debugger API designed primarily for
+writing comprehensive unit tests for the eBPF verifier. It allows developers
+to introspect the abstract interpretation process step-by-step to guarantee
+that the verifier correctly models the semantics of eBPF instructions.
+
+The debugger operates using a gdb-like approach:
+
+1. **Initialization:** Create a debug session using
+ ``rte_bpf_validate_debug_create()`` and pass it to the loader via the
+ ``debug`` field in ``struct rte_bpf_prm_ex``.
+2. **Breakpoints and Catchpoints:** Before loading, use
+ ``rte_bpf_validate_debug_break()`` or ``rte_bpf_validate_debug_catch()``
+ to register callback functions that trigger at specific instruction indices
+ (program counters) or upon specific validation events.
+3. **State Introspection:** Within the callbacks, the API provides functions
+ like ``rte_bpf_validate_debug_can_access()``,
+ ``rte_bpf_validate_debug_may_jump()``, and various formatting functions
+ to safely inspect the verifier's internal belief about register bounds
+ and memory states at that specific execution point.
+
+When adding a test for a new eBPF instruction or fixing a validator bug,
+developers should utilize the harness provided in
+``app/test/test_bpf_validate.c``. This harness encapsulates the debugger API,
+allowing you to define the expected abstract domains (signed and unsigned
+intervals) for registers before and after a tested instruction, generating
+the necessary eBPF bytecode and breakpoints automatically.
+
+
Not currently supported eBPF features
-------------------------------------
--
2.43.0
^ permalink raw reply related
* [PATCH v3 24/25] doc: add release notes for BPF validation fixes
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
Cc: dev, Konstantin Ananyev
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Document the following new features and fixes:
* Added BPF validation debugger API (rte_bpf_validate_debug_*).
* Hardened BPF validator with numerous bug fixes and UB preventions.
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
doc/guides/rel_notes/release_26_07.rst | 16 ++++++++++++++++
1 file changed, 16 insertions(+)
diff --git a/doc/guides/rel_notes/release_26_07.rst b/doc/guides/rel_notes/release_26_07.rst
index 3a151c8e83bb..d98e73b8340f 100644
--- a/doc/guides/rel_notes/release_26_07.rst
+++ b/doc/guides/rel_notes/release_26_07.rst
@@ -175,6 +175,22 @@ New Features
``rte_bpf_eth_tx_install`` for installing already loaded BPF programs as
port callbacks (as opposed to loading them directly from ELF files).
+* **Hardened BPF validator.**
+
+ Fixed numerous bugs in the BPF validator's abstract interpretation logic,
+ including incorrect bounds tracking for jumps and arithmetic operations, as
+ well as fixing several instances of undefined behavior (UB) when verifying
+ malicious or corrupt programs.
+
+* **Added BPF validation debugger API.**
+
+ Introduced a new set of APIs (prefixed with ``rte_bpf_validate_debug_``) to
+ introspect the BPF validator. This provides a mechanism to set breakpoints or
+ catchpoints during validation and inspect the verifier's internal state
+ (such as tracked register bounds). This API is crucial primarily for writing
+ comprehensive tests for the validator, but also serves as a foundation for a
+ future interactive eBPF validation debugger.
+
Removed Items
-------------
--
2.43.0
^ permalink raw reply related
* [PATCH v3 23/25] bpf/validate: prevent overflow when building graph
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Function `evst_pool_init` for malicious or corrupt BPF program with
number of conditional jumps exceeding a third of UINT32_MAX could cause
arithmetic and buffer overflows when working with the program graph.
Fix the issue by limiting maximum number of conditional jumps supported
by UINT32_MAX / 4, or more than 1 billion.
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
lib/bpf/bpf_validate.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index 03c590c75377..f9960088a285 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -2662,6 +2662,10 @@ evst_pool_init(struct bpf_verifier *bvf)
{
uint32_t k, n;
+ if (bvf->nb_jcc_nodes > UINT32_MAX / 4)
+ /* Calculations that follow may overflow. */
+ return -E2BIG;
+
/*
* We need nb_jcc_nodes + 1 for save_cur/restore_cur
* remaining ones will be used for state tracking/pruning.
--
2.43.0
^ permalink raw reply related
* [PATCH v3 22/25] bpf/validate: fix BPF_XOR signed min calculation
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Function `eval_xor` calculated signed minimum using essentially unsigned
algorithm as long as any of the operands have non-negative range, which
is incorrect since it ignores any negative numbers that may have the
sign or any other bits set.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: ldxdw r2, [r1 + 0]
2: jsgt r2, #0x0, L5
3: xor r2, #0x0 ; tested instruction
4: mov r0, #0x1
5: exit
Pre-state:
r2: INT64_MIN..0
Post-state:
r2: 0
After the tested instruction validator considers r2 to equal 0, however
if -1 was loaded on step 1 it is possible for it to be -1.
Set signed range to full if any of the operands can be negative,
otherwise (if both operands are non-negative) use same algorithm as for
unsigned numbers. Add test.
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
app/test/test_bpf_validate.c | 17 +++++++++++++++++
lib/bpf/bpf_validate.c | 2 +-
2 files changed, 18 insertions(+), 1 deletion(-)
diff --git a/app/test/test_bpf_validate.c b/app/test/test_bpf_validate.c
index 205373a4f86b..a06d3254d6ba 100644
--- a/app/test/test_bpf_validate.c
+++ b/app/test/test_bpf_validate.c
@@ -1764,6 +1764,23 @@ test_alu64_sub_x_src_signed_max_zero(void)
REGISTER_FAST_TEST(bpf_validate_alu64_sub_x_src_signed_max_zero_autotest, NOHUGE_OK, ASAN_OK,
test_alu64_sub_x_src_signed_max_zero);
+/* 64-bit bitwise XOR between a negative scalar range and zero immediate. */
+static int
+test_alu64_xor_k_negative(void)
+{
+ return verify_instruction((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (EBPF_ALU64 | BPF_XOR | BPF_K),
+ .imm = 0,
+ },
+ .pre.dst = make_signed_domain(INT64_MIN, 0),
+ .post.dst = unknown,
+ });
+}
+
+REGISTER_FAST_TEST(bpf_validate_alu64_xor_k_negative_autotest, NOHUGE_OK, ASAN_OK,
+ test_alu64_xor_k_negative);
+
/* Jump if greater than immediate. */
static int
test_jmp64_jeq_k(void)
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index 131a5468dbc4..03c590c75377 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -910,7 +910,7 @@ eval_xor(struct bpf_reg_val *rd, const struct bpf_reg_val *rs, size_t opsz,
rd->s.max ^= rs->s.max;
/* both operands are non-negative */
- } else if (rd->s.min >= 0 || rs->s.min >= 0) {
+ } else if (rd->s.min >= 0 && rs->s.min >= 0) {
rd->s.max = eval_uor_max(rd->s.max, rs->s.max, opsz);
rd->s.min = 0;
} else
--
2.43.0
^ permalink raw reply related
* [PATCH v3 21/25] bpf/validate: fix BPF_SUB signed max zero case
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Function `eval_sub` used source register signed minimum to detect
overflow of the difference (operation result) signed minimum, and source
register signed maximum to detect overflow of the difference signed
maximum. However in the actual formula for difference source register
bounds are swapped (correctly, since we subtract it), so in overflow
detection we should also have swapped them. It caused false negatives in
certain cases.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: ldxdw r2, [r1 + 0]
2: jsgt r2, #0x0, L7
3: ldxdw r3, [r1 + 8]
4: jsgt r3, #0x0, L7
5: sub r2, r3 ; tested instruction
6: mov r0, #0x1
7: exit
Pre-state:
r2: INT64_MIN..0
r3: INT64_MIN..0
Post-state:
r2: INT64_MIN
Validator ignores overflow of signed minimum and considers result to
always equal INT64_MIN. However, if -1 was loaded on step 1 and -2 was
loaded on step 3 it is possible for the difference to equal 1.
Swap source register signed minimum and maximum in the overflow
condition to match the new range formula, add test.
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
app/test/test_bpf_validate.c | 17 +++++++++++++++++
lib/bpf/bpf_validate.c | 4 ++--
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/app/test/test_bpf_validate.c b/app/test/test_bpf_validate.c
index 7cf9e404b697..205373a4f86b 100644
--- a/app/test/test_bpf_validate.c
+++ b/app/test/test_bpf_validate.c
@@ -1747,6 +1747,23 @@ test_alu64_or_k_positive(void)
REGISTER_FAST_TEST(bpf_validate_alu64_or_k_positive_autotest, NOHUGE_OK, ASAN_OK,
test_alu64_or_k_positive);
+/* 64-bit difference between two negative ranges.. */
+static int
+test_alu64_sub_x_src_signed_max_zero(void)
+{
+ return verify_instruction((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (EBPF_ALU64 | BPF_SUB | BPF_X),
+ },
+ .pre.dst = make_signed_domain(INT64_MIN, 0),
+ .pre.src = make_signed_domain(INT64_MIN, 0),
+ .post.dst = unknown,
+ });
+}
+
+REGISTER_FAST_TEST(bpf_validate_alu64_sub_x_src_signed_max_zero_autotest, NOHUGE_OK, ASAN_OK,
+ test_alu64_sub_x_src_signed_max_zero);
+
/* Jump if greater than immediate. */
static int
test_jmp64_jeq_k(void)
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index abb39cfd328d..131a5468dbc4 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -716,9 +716,9 @@ eval_sub(struct bpf_reg_val *rd, const struct bpf_reg_val *rs, uint64_t msk)
eval_umax_bound(&rv, msk);
if ((rd->s.min != rd->s.max || rs->s.min != rs->s.max) &&
- (((rs->s.min < 0 && rv.s.min < rd->s.min) ||
+ (((rs->s.max < 0 && rv.s.min < rd->s.min) ||
rv.s.min > rd->s.min) ||
- ((rs->s.max < 0 && rv.s.max < rd->s.max) ||
+ ((rs->s.min < 0 && rv.s.max < rd->s.max) ||
rv.s.max > rd->s.max)))
eval_smax_bound(&rv, msk);
--
2.43.0
^ permalink raw reply related
* [PATCH v3 20/25] bpf/validate: fix BPF_OR min calculations
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
This commit fixes two different problems in signed and unsigned minimum
calculations within `eval_or`. Passing tests requires both problems to
be fixed which is why the changes are squashed in one commit.
1) Function `eval_or` calculated result signed minimum as bitwise OR
between corresponding minimums as long as any of them is non-negative,
which is incorrect since values within the range can have zeroes where
the minimums don't, including the sign bit.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: ldxdw r2, [r1 + 0]
2: jlt r2, #0x5, L8
3: jgt r2, #0x6, L8
4: jslt r2, #0x5, L8
5: jsgt r2, #0x6, L8
6: or r2, #0xfffffffe ; tested instruction
7: mov r0, #0x1
8: exit
Pre-state:
r2: 5..6
Post-state:
r2: -1
After the tested instruction validator considers r2 to always equal -1,
however if 6 was loaded on step 1 it is possible for it to be -2:
0x6 & 0xfffffffffffffffe == 0xfffffffffffffffe = -2
Set signed range to full if any of the operands can be negative,
otherwise use the maximum of both minimums as a new signed minimum
following the idea that result of bitwise OR cannot be smaller than its
operands. Add test.
2) Function `eval_or` calculated result unsigned minimum as bitwise OR
between corresponding minimums, which is incorrect since values within
the range can have zeroes the minimums don't.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: ldxdw r2, [r1 + 0]
2: jlt r2, #0x5, L8
3: jgt r2, #0x6, L8
4: jslt r2, #0x5, L8
5: jsgt r2, #0x6, L8
6: or r2, #0x2 ; tested instruction
7: mov r0, #0x1
8: exit
Pre-state:
r2: 5..6
Post-state:
r2: 7
After the tested instruction validator considers r2 to always equal 7,
however if 6 was loaded on step 1 it is possible for it to be 6:
0x6 & 0x2 == 0x6
Use the maximum of both minimums as a new unsigned minimum following the
idea that result of bitwise OR cannot be smaller than its operands. Add
test.
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
app/test/test_bpf_validate.c | 34 ++++++++++++++++++++++++++++++++++
lib/bpf/bpf_validate.c | 6 +++---
2 files changed, 37 insertions(+), 3 deletions(-)
diff --git a/app/test/test_bpf_validate.c b/app/test/test_bpf_validate.c
index c3d1bdb78fbc..7cf9e404b697 100644
--- a/app/test/test_bpf_validate.c
+++ b/app/test/test_bpf_validate.c
@@ -1713,6 +1713,40 @@ test_alu64_neg_zero_last(void)
REGISTER_FAST_TEST(bpf_validate_alu64_neg_zero_last_autotest, NOHUGE_OK, ASAN_OK,
test_alu64_neg_zero_last);
+/* 64-bit bitwise OR between a positive scalar range and negative immediate. */
+static int
+test_alu64_or_k_negative(void)
+{
+ return verify_instruction((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (EBPF_ALU64 | BPF_OR | BPF_K),
+ .imm = -2,
+ },
+ .pre.dst = make_signed_domain(5, 6),
+ .post.dst = make_signed_domain(-2, -1),
+ });
+}
+
+REGISTER_FAST_TEST(bpf_validate_alu64_or_k_negative_autotest, NOHUGE_OK, ASAN_OK,
+ test_alu64_or_k_negative);
+
+/* 64-bit bitwise OR between a positive scalar range and positive immediate. */
+static int
+test_alu64_or_k_positive(void)
+{
+ return verify_instruction((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (EBPF_ALU64 | BPF_OR | BPF_K),
+ .imm = 2,
+ },
+ .pre.dst = make_signed_domain(5, 6),
+ .post.dst = make_signed_domain(5, 7),
+ });
+}
+
+REGISTER_FAST_TEST(bpf_validate_alu64_or_k_positive_autotest, NOHUGE_OK, ASAN_OK,
+ test_alu64_or_k_positive);
+
/* Jump if greater than immediate. */
static int
test_jmp64_jeq_k(void)
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index 4e4c0ddeb2b8..abb39cfd328d 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -875,7 +875,7 @@ eval_or(struct bpf_reg_val *rd, const struct bpf_reg_val *rs, size_t opsz,
rd->u.max |= rs->u.max;
} else {
rd->u.max = eval_uor_max(rd->u.max, rs->u.max, opsz);
- rd->u.min |= rs->u.min;
+ rd->u.min = RTE_MAX(rd->u.min, rs->u.min);
}
/* both operands are constants */
@@ -884,9 +884,9 @@ eval_or(struct bpf_reg_val *rd, const struct bpf_reg_val *rs, size_t opsz,
rd->s.max |= rs->s.max;
/* both operands are non-negative */
- } else if (rd->s.min >= 0 || rs->s.min >= 0) {
+ } else if (rd->s.min >= 0 && rs->s.min >= 0) {
rd->s.max = eval_uor_max(rd->s.max, rs->s.max, opsz);
- rd->s.min |= rs->s.min;
+ rd->s.min = RTE_MAX(rd->s.min, rs->s.min);
} else
eval_smax_bound(rd, msk);
}
--
2.43.0
^ permalink raw reply related
* [PATCH v3 19/25] bpf/validate: fix BPF_LSH shift-out-of-bounds UB
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Function `eval_lsh` when validating left shift by 63 invoked macro
`RTE_LEN2MASK(0, int64_t)` which triggered shift-out-of-bounds undefined
behaviour.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: ldxdw r2, [r1 + 0]
2: jlt r2, #0x3, L8
3: jgt r2, #0x5, L8
4: jslt r2, #0x3, L8
5: jsgt r2, #0x5, L8
6: lsh r2, #0x3f ; tested instruction
7: mov r0, #0x1
8: exit
Pre-state:
r2: 3..5
Post-state:
r2: 0..UINT64_MAX
With sanitizer the following diagnostic is generated:
lib/bpf/bpf_validate.c:785:4: runtime error: shift exponent 64 is
too large for 64-bit type 'long unsigned int'
#0 0x00000274d5e0 in eval_lsh lib/bpf/bpf_validate.c:785
#1 0x00000275a2ea in eval_alu lib/bpf/bpf_validate.c:1310
#2 0x00000276ce3d in evaluate lib/bpf/bpf_validate.c:3284
Add guard for this case, add test.
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
app/test/test_bpf_validate.c | 17 +++++++++++++++++
lib/bpf/bpf_validate.c | 3 ++-
2 files changed, 19 insertions(+), 1 deletion(-)
diff --git a/app/test/test_bpf_validate.c b/app/test/test_bpf_validate.c
index 5d26299ba65d..c3d1bdb78fbc 100644
--- a/app/test/test_bpf_validate.c
+++ b/app/test/test_bpf_validate.c
@@ -1536,6 +1536,23 @@ test_alu64_div_mod_overflow(void)
REGISTER_FAST_TEST(bpf_validate_alu64_div_mod_overflow_autotest, NOHUGE_OK, ASAN_OK,
test_alu64_div_mod_overflow);
+/* 64-bit left shift by 63. */
+static int
+test_alu64_lsh_63(void)
+{
+ return verify_instruction((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (EBPF_ALU64 | BPF_LSH | BPF_K),
+ .imm = 63,
+ },
+ .pre.dst = make_signed_domain(3, 5),
+ .post.dst = unknown,
+ });
+}
+
+REGISTER_FAST_TEST(bpf_validate_alu64_lsh_63_autotest, NOHUGE_OK, ASAN_OK,
+ test_alu64_lsh_63);
+
/* 64-bit multiplication of constant and immediate with overflow. */
static int
test_alu64_mul_k_overflow(void)
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index d4d8ec4251f1..4e4c0ddeb2b8 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -746,7 +746,8 @@ eval_lsh(struct bpf_reg_val *rd, const struct bpf_reg_val *rs, size_t opsz,
/* check that dreg values are and would remain always positive */
if ((uint64_t)rd->s.min >> (opsz - 1) != 0 || rd->s.max >=
- RTE_LEN2MASK(opsz - rs->u.max - 1, int64_t))
+ (rs->u.max == opsz - 1 ? 0 :
+ RTE_LEN2MASK(opsz - rs->u.max - 1, int64_t)))
eval_smax_bound(rd, msk);
else {
rd->s.max <<= rs->u.max;
--
2.43.0
^ permalink raw reply related
* [PATCH v3 18/25] bpf/validate: fix BPF_AND min calculations
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Function `eval_and` calculated both signed (if positive) and unsigned
minimum values as bitwise AND between corresponding minimums, which is
incorrect since intermediate values can have zeroes in bits where
minimum values don't.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: ldxdw r2, [r1 + 0]
2: jlt r2, #0x6, L8
3: jgt r2, #0x8, L8
4: jslt r2, #0x6, L8
5: jsgt r2, #0x8, L8
6: and r2, #0x5 ; tested instruction
7: mov r0, #0x1
8: exit
Pre-state:
r2: 6..8
Post-state:
r2: 4..7
After the tested instruction validator considers r2 to be equal or
greater than 4, however if 8 was loaded on step 1 it is possible for it
to be zero (0x8 & 0x5 == 0).
Use zero as a new safe lower bound for both signed (if positive) and
unsigned minimum. Add test.
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
app/test/test_bpf_validate.c | 17 +++++++++++++++++
lib/bpf/bpf_validate.c | 4 ++--
2 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/app/test/test_bpf_validate.c b/app/test/test_bpf_validate.c
index f80dee24a677..5d26299ba65d 100644
--- a/app/test/test_bpf_validate.c
+++ b/app/test/test_bpf_validate.c
@@ -1384,6 +1384,23 @@ test_alu64_add_x_scalar_scalar(void)
REGISTER_FAST_TEST(bpf_validate_alu64_add_x_scalar_scalar_autotest, NOHUGE_OK, ASAN_OK,
test_alu64_add_x_scalar_scalar);
+/* 64-bit bitwise AND between a scalar range and immediate. */
+static int
+test_alu64_and_k(void)
+{
+ return verify_instruction((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (EBPF_ALU64 | BPF_AND | BPF_K),
+ .imm = 5,
+ },
+ .pre.dst = make_signed_domain(6, 8),
+ .post.dst = make_signed_domain(0, 7),
+ });
+}
+
+REGISTER_FAST_TEST(bpf_validate_alu64_and_k_autotest, NOHUGE_OK, ASAN_OK,
+ test_alu64_and_k);
+
/* 64-bit division and modulo of UINT64_MAX*2/3. */
static int
test_alu64_div_mod_big_constant(void)
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index af084e36c8d0..d4d8ec4251f1 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -848,7 +848,7 @@ eval_and(struct bpf_reg_val *rd, const struct bpf_reg_val *rs, size_t opsz,
rd->u.max &= rs->u.max;
} else {
rd->u.max = eval_uand_max(rd->u.max, rs->u.max, opsz);
- rd->u.min &= rs->u.min;
+ rd->u.min = 0;
}
/* both operands are constants */
@@ -859,7 +859,7 @@ eval_and(struct bpf_reg_val *rd, const struct bpf_reg_val *rs, size_t opsz,
} else if (rd->s.min >= 0 || rs->s.min >= 0) {
rd->s.max = eval_uand_max(rd->s.max & (msk >> 1),
rs->s.max & (msk >> 1), opsz);
- rd->s.min &= rs->s.min;
+ rd->s.min = 0;
} else
eval_smax_bound(rd, msk);
}
--
2.43.0
^ permalink raw reply related
* [PATCH v3 17/25] bpf/validate: fix BPF_JMP empty range handling
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
Function `eval_jcc` did not account for 'dynamically unreachable' code
paths. Some code paths may be _dynamically_ unreachable, which measn
that according to validator calculations no valid values are left to
evaluate. This does not indicate dead code since same code might be
reachable through other code paths. Previous behaviour resulted in:
* undefined behaviour in corner cases;
* ranges breaking min <= max invariant relied upon in multiple places
(e.g. signed overflow detection in `eval_mul` only checks `s.min` to
make sure the range is non-negative and so on);
* unnecessary work for validator contributing to exponential code paths
grow in some cases.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: mov r2, #0x2a
2: lddw r3, #0x8000000000000000
4: jslt r2, r3, L7 ; tested instruction
5: mov r0, #0x1
6: exit
7: mov r0, #0x2
8: exit
Pre-state:
r2: 42
r3: INT64_MIN
Post-state:
r2: 42
r3: INT64_MIN
Jump-state:
r2: 42
r3: 43..INT64_MIN INTERSECT 0x8000000000000000 (!)
At step 7 after jump from tested instruction validator considers r3 to
equal 0x8000000000000000 if viewed as unsigned, or have nonsensical
range 43..INT64_MIN if viewed as signed. In reality there is just no
valid range for this code path since it will never occur.
With sanitizer the following diagnostic is generated:
lib/bpf/bpf_validate.c:1824:15: runtime error: signed integer
overflow: -9223372036854775808 - 1 cannot be represented in type
'long int'
#0 0x000002761e41 in eval_jslt_jsge lib/bpf/bpf_validate.c:1824
#1 0x000002762acb in eval_jcc lib/bpf/bpf_validate.c:1881
#2 0x00000276b749 in evaluate lib/bpf/bpf_validate.c:3245
...
SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior
lib/bpf/bpf_validate.c:1824:15
Add pruning of dynamically unreachable code paths that arise from
ordering comparisons. Add tests for remaining ordering jump cases.
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
app/test/test_bpf_validate.c | 277 ++++++++++++++++++++++++++++++-
lib/bpf/bpf_validate.c | 96 ++++++++---
lib/bpf/rte_bpf_validate_debug.h | 2 +
3 files changed, 351 insertions(+), 24 deletions(-)
diff --git a/app/test/test_bpf_validate.c b/app/test/test_bpf_validate.c
index 15ccf4f6a573..f80dee24a677 100644
--- a/app/test/test_bpf_validate.c
+++ b/app/test/test_bpf_validate.c
@@ -135,6 +135,11 @@ static const struct domain unknown = {
.u = { .min = 0, .max = UINT64_MAX },
};
+/* Unreachable state. */
+static const struct state unreachable = {
+ .is_unreachable = true,
+};
+
/* BUILDING DOMAINS */
@@ -1710,6 +1715,55 @@ test_jmp64_jslt_x(void)
REGISTER_FAST_TEST(bpf_validate_jmp64_jslt_x_autotest, NOHUGE_OK, ASAN_OK,
test_jmp64_jslt_x);
+/* Jump on ordering comparisons with potential bound overflow. */
+static int
+test_jmp64_ordering_overflow(void)
+{
+ /* In this test signed and unsigned cases are spelled out explicitly. */
+ const bool also_signed = false;
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JSLT | BPF_X),
+ },
+ .pre.dst = make_singleton_domain(42),
+ .pre.src = make_singleton_domain(INT64_MIN),
+ .jump = unreachable,
+ }, also_signed), "signed less than INT64_MIN");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JSGT | BPF_X),
+ },
+ .pre.dst = make_singleton_domain(42),
+ .pre.src = make_singleton_domain(INT64_MAX),
+ .jump = unreachable,
+ }, also_signed), "signed greater than INT64_MAX");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
+ },
+ .pre.dst = make_singleton_domain(42),
+ .pre.src = make_singleton_domain(0),
+ .jump = unreachable,
+ }, also_signed), "unsigned less than zero");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | BPF_JGT | BPF_X),
+ },
+ .pre.dst = make_singleton_domain(42),
+ .pre.src = make_singleton_domain(UINT64_MAX),
+ .jump = unreachable,
+ }, also_signed), "unsigned greater than UINT64_MAX");
+
+ return TEST_SUCCESS;
+}
+
+REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_overflow_autotest, NOHUGE_OK, ASAN_OK,
+ test_jmp64_ordering_overflow);
+
/* Jump on ordering comparisons between two ranges. */
static int
test_jmp64_ordering_ranges(void)
@@ -1717,6 +1771,29 @@ test_jmp64_ordering_ranges(void)
/* All ranges used are valid for both signed and unsigned comparisons. */
const bool also_signed = true;
+ /*
+ * 20 ---- dst ---- 60
+ * 0 - src - 10
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .pre.src = make_signed_domain(0, 10),
+ .jump = unreachable,
+ }, also_signed), "strict, dst range strongly greater than src range");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .pre.src = make_signed_domain(0, 10),
+ .jump = unreachable,
+ }, also_signed), "non-strict, dst range strongly greater than src range");
+
/*
* 20 ---- dst ---- 60
* 10 -- src -- 40
@@ -1817,15 +1894,38 @@ test_jmp64_ordering_ranges(void)
.post.src = make_signed_domain(40, 59),
}, also_signed), "non-strict, dst range weakly less than src range");
+ /*
+ * 20 ---- dst ---- 60
+ * 70 - src - 80
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .pre.src = make_signed_domain(70, 80),
+ .post = unreachable,
+ }, also_signed), "strict, dst range strongly less than src range");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .pre.src = make_signed_domain(70, 80),
+ .post = unreachable,
+ }, also_signed), "non-strict, dst range strongly less than src range");
+
return TEST_SUCCESS;
}
REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_ranges_autotest, NOHUGE_OK, ASAN_OK,
test_jmp64_ordering_ranges);
-/* Jump on ordering comparisons with singleton. */
+/* Jump on ordering comparisons with singleton inside the range. */
static int
-test_jmp64_ordering_singleton(void)
+test_jmp64_ordering_singleton_inside(void)
{
/* All ranges used are valid for both signed and unsigned comparisons. */
const bool also_signed = true;
@@ -1878,8 +1978,177 @@ test_jmp64_ordering_singleton(void)
return TEST_SUCCESS;
}
-REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_singleton_autotest, NOHUGE_OK, ASAN_OK,
- test_jmp64_ordering_singleton);
+REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_singleton_inside_autotest, NOHUGE_OK, ASAN_OK,
+ test_jmp64_ordering_singleton_inside);
+
+/* Jump on ordering comparisons with singleton outside the range. */
+static int
+test_jmp64_ordering_singleton_outside(void)
+{
+ /* All ranges used are valid for both signed and unsigned comparisons. */
+ const bool also_signed = true;
+
+ /*
+ * 20 ---- dst ---- 60
+ * imm
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_K),
+ .imm = 10,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .jump = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JLT | BPF_K) check, range greater than imm");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_K),
+ .imm = 10,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .jump = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JLE | BPF_K) check, range greater than imm");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | BPF_JGT | BPF_K),
+ .imm = 10,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .post = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JGT | BPF_K) check, range greater than imm");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | BPF_JGE | BPF_K),
+ .imm = 10,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .post = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JGE | BPF_K) check, range greater than imm");
+
+ /*
+ * 20 ---- dst ---- 60
+ * imm
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_K),
+ .imm = 70,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .post = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JLT | BPF_K) check, range less than imm");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_K),
+ .imm = 70,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .post = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JLE | BPF_K) check, range less than imm");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | BPF_JGT | BPF_K),
+ .imm = 70,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .jump = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JGT | BPF_K) check, range less than imm");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | BPF_JGE | BPF_K),
+ .imm = 70,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .jump = unreachable,
+ }, also_signed), "(BPF_JMP | EBPF_JGE | BPF_K) check, range less than imm");
+
+ return TEST_SUCCESS;
+}
+
+REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_singleton_outside_autotest, NOHUGE_OK, ASAN_OK,
+ test_jmp64_ordering_singleton_outside);
+
+/* Jump on ordering comparisons with ranges "touching" each other. */
+static int
+test_jmp64_ordering_touching(void)
+{
+ /* All ranges used are valid for both signed and unsigned comparisons. */
+ const bool also_signed = true;
+
+ for (int overlap = 0; overlap != 3; ++overlap) {
+
+ /*
+ * 20 - dst - 30
+ * 10 - src - (19 + overlap)
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 30),
+ .pre.src = make_signed_domain(10, 19 + overlap),
+ .jump = overlap <= 1 ? unreachable : (struct state){
+ .dst = make_singleton_domain(20),
+ .src = make_singleton_domain(21),
+ },
+ }, also_signed), "strict, dst left touching src right, overlap=%d", overlap);
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 30),
+ .pre.src = make_signed_domain(10, 19 + overlap),
+ .jump = overlap < 1 ? unreachable : (struct state){
+ .dst = make_signed_domain(20, 19 + overlap),
+ .src = make_signed_domain(20, 19 + overlap),
+ },
+ }, also_signed), "non-strict, dst left touching src right, overlap=%d", overlap);
+
+ /*
+ * 10 - dst - (19 + overlap)
+ * 20 - src - 30
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
+ },
+ .pre.dst = make_signed_domain(10, 19 + overlap),
+ .pre.src = make_signed_domain(20, 30),
+ .post = overlap < 1 ? unreachable : (struct state){
+ .dst = make_signed_domain(20, 19 + overlap),
+ .src = make_signed_domain(20, 19 + overlap),
+ },
+ }, also_signed), "strict, dst right touching src left, overlap=%d", overlap);
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_X),
+ },
+ .pre.dst = make_signed_domain(10, 19 + overlap),
+ .pre.src = make_signed_domain(20, 30),
+ .post = overlap <= 1 ? unreachable : (struct state){
+ .dst = make_singleton_domain(21),
+ .src = make_singleton_domain(20),
+ },
+ }, also_signed), "non-strict, dst right touching src left, overlap=%d", overlap);
+ }
+
+ return TEST_SUCCESS;
+}
+
+REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_touching_autotest, NOHUGE_OK, ASAN_OK,
+ test_jmp64_ordering_touching);
/* 64-bit load from heap (should be set to unknown). */
static int
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index 2e535069fe4d..af084e36c8d0 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -19,6 +19,9 @@
#define BPF_ARG_PTR_STACK RTE_BPF_ARG_RESERVED
+/* type containing no values (AKA "bottom", "never" etc) */
+#define BPF_ARG_UNINHABITED ((enum rte_bpf_arg_type)(RTE_BPF_ARG_UNDEF - 1))
+
struct bpf_reg_val {
struct rte_bpf_arg v;
uint64_t mask;
@@ -36,6 +39,8 @@ struct bpf_eval_state {
SLIST_ENTRY(bpf_eval_state) next; /* for @safe list traversal */
struct bpf_reg_val rv[EBPF_REG_NUM];
struct bpf_reg_val sv[MAX_BPF_STACK_SIZE / sizeof(uint64_t)];
+ /* flag set for branches determined to be dynamically unreachable */
+ bool unreachable;
};
SLIST_HEAD(bpf_evst_head, bpf_eval_state);
@@ -174,6 +179,9 @@ __rte_bpf_validate_can_access(const struct bpf_verifier *verifier,
struct value_set access_set;
uint32_t opsz;
+ if (st->unreachable)
+ return -ENOENT;
+
switch (BPF_CLASS(access->code)) {
case BPF_LDX:
rv = &st->rv[access->src_reg];
@@ -310,6 +318,10 @@ __rte_bpf_validate_may_jump(const struct bpf_verifier *verifier,
if (!may_jump_code_is_supported(jump->code))
return -ENOTSUP;
+ if (st->unreachable)
+ /* Set no bits since neither false nor true is possible. */
+ return 0;
+
rd = &st->rv[jump->dst_reg];
dst_set = (rd->v.type == RTE_BPF_ARG_UNDEF) ? value_set_full :
value_set_from_pair(rd->s.min, rd->s.max, rd->u.min, rd->u.max);
@@ -1521,40 +1533,68 @@ static void
eval_jgt_jle(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
- frd->u.max = RTE_MIN(frd->u.max, frs->u.max);
- frs->u.min = RTE_MAX(frs->u.min, frd->u.min);
- trd->u.min = RTE_MAX(trd->u.min, trs->u.min + 1);
- trs->u.max = RTE_MIN(trs->u.max, trd->u.max - 1);
+ if (frd->u.min <= frs->u.max) {
+ frd->u.max = RTE_MIN(frd->u.max, frs->u.max);
+ frs->u.min = RTE_MAX(frs->u.min, frd->u.min);
+ } else
+ frd->v.type = frs->v.type = BPF_ARG_UNINHABITED;
+
+ if (trs->u.min < trd->u.max) {
+ trd->u.min = RTE_MAX(trd->u.min, trs->u.min + 1);
+ trs->u.max = RTE_MIN(trs->u.max, trd->u.max - 1);
+ } else
+ trd->v.type = trs->v.type = BPF_ARG_UNINHABITED;
}
static void
eval_jlt_jge(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
- frd->u.min = RTE_MAX(frd->u.min, frs->u.min);
- frs->u.max = RTE_MIN(frs->u.max, frd->u.max);
- trd->u.max = RTE_MIN(trd->u.max, trs->u.max - 1);
- trs->u.min = RTE_MAX(trs->u.min, trd->u.min + 1);
+ if (frs->u.min <= frd->u.max) {
+ frd->u.min = RTE_MAX(frd->u.min, frs->u.min);
+ frs->u.max = RTE_MIN(frs->u.max, frd->u.max);
+ } else
+ frd->v.type = frs->v.type = BPF_ARG_UNINHABITED;
+
+ if (trd->u.min < trs->u.max) {
+ trd->u.max = RTE_MIN(trd->u.max, trs->u.max - 1);
+ trs->u.min = RTE_MAX(trs->u.min, trd->u.min + 1);
+ } else
+ trd->v.type = trs->v.type = BPF_ARG_UNINHABITED;
}
static void
eval_jsgt_jsle(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
- frd->s.max = RTE_MIN(frd->s.max, frs->s.max);
- frs->s.min = RTE_MAX(frs->s.min, frd->s.min);
- trd->s.min = RTE_MAX(trd->s.min, trs->s.min + 1);
- trs->s.max = RTE_MIN(trs->s.max, trd->s.max - 1);
+ if (frd->s.min <= frs->s.max) {
+ frd->s.max = RTE_MIN(frd->s.max, frs->s.max);
+ frs->s.min = RTE_MAX(frs->s.min, frd->s.min);
+ } else
+ frd->v.type = frs->v.type = BPF_ARG_UNINHABITED;
+
+ if (trs->s.min < trd->s.max) {
+ trd->s.min = RTE_MAX(trd->s.min, trs->s.min + 1);
+ trs->s.max = RTE_MIN(trs->s.max, trd->s.max - 1);
+ } else
+ trd->v.type = trs->v.type = BPF_ARG_UNINHABITED;
}
static void
eval_jslt_jsge(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
- frd->s.min = RTE_MAX(frd->s.min, frs->s.min);
- frs->s.max = RTE_MIN(frs->s.max, frd->s.max);
- trd->s.max = RTE_MIN(trd->s.max, trs->s.max - 1);
- trs->s.min = RTE_MAX(trs->s.min, trd->s.min + 1);
+ if (frs->s.min <= frd->s.max) {
+ frd->s.min = RTE_MAX(frd->s.min, frs->s.min);
+ frs->s.max = RTE_MIN(frs->s.max, frd->s.max);
+ } else
+ frd->v.type = frs->v.type = BPF_ARG_UNINHABITED;
+
+ if (trd->s.min < trs->s.max) {
+ trd->s.max = RTE_MIN(trd->s.max, trs->s.max - 1);
+ trs->s.min = RTE_MAX(trs->s.min, trd->s.min + 1);
+ } else
+ trd->v.type = trs->v.type = BPF_ARG_UNINHABITED;
}
static const char *
@@ -1609,6 +1649,14 @@ eval_jcc(struct bpf_verifier *bvf, const struct ebpf_insn *ins)
else if (op == EBPF_JSGE)
eval_jslt_jsge(frd, frs, trd, trs);
+ if (trd->v.type == BPF_ARG_UNINHABITED ||
+ trs->v.type == BPF_ARG_UNINHABITED)
+ tst->unreachable = true;
+
+ if (frd->v.type == BPF_ARG_UNINHABITED ||
+ frs->v.type == BPF_ARG_UNINHABITED)
+ fst->unreachable = true;
+
return NULL;
}
@@ -2349,7 +2397,7 @@ set_edge_type(struct bpf_verifier *bvf, struct inst_node *node,
* Depth-First Search (DFS) through previously constructed
* Control Flow Graph (CFG).
* Information collected at this path would be used later
- * to determine is there any loops, and/or unreachable instructions.
+ * to determine is there any loops, and/or statically unreachable instructions.
* PREREQUISITE: there is at least one node.
*/
static void
@@ -2397,7 +2445,7 @@ dfs(struct bpf_verifier *bvf)
}
/*
- * report unreachable instructions.
+ * report statically unreachable instructions.
*/
static void
log_unreachable(const struct bpf_verifier *bvf)
@@ -2970,13 +3018,21 @@ evaluate(struct bpf_verifier *bvf)
stats.nb_restore++;
}
+ if (bvf->evst->unreachable) {
+ rc = __rte_bpf_validate_debug_evaluate_step(
+ debug, get_node_idx(bvf, next),
+ RTE_BPF_VALIDATE_DEBUG_EVENT_BRANCH_UNREACHABLE);
+ if (rc < 0)
+ break;
+
+ next = NULL;
/*
* for jcc targets: check did we already evaluated
* that path and can it's evaluation be skipped that
* time.
*/
- if (node->nb_edge > 1 && prune_eval_state(bvf, node,
- next) == 0) {
+ } else if (node->nb_edge > 1 &&
+ prune_eval_state(bvf, node, next) == 0) {
rc = __rte_bpf_validate_debug_evaluate_step(
debug, get_node_idx(bvf, next),
RTE_BPF_VALIDATE_DEBUG_EVENT_BRANCH_PRUNE);
diff --git a/lib/bpf/rte_bpf_validate_debug.h b/lib/bpf/rte_bpf_validate_debug.h
index 89bf587f0211..f30fa926f10a 100644
--- a/lib/bpf/rte_bpf_validate_debug.h
+++ b/lib/bpf/rte_bpf_validate_debug.h
@@ -49,6 +49,8 @@ enum rte_bpf_validate_debug_event {
RTE_BPF_VALIDATE_DEBUG_EVENT_BRANCH_PRUNE,
/* End of branch verification, after the last verified instruction. */
RTE_BPF_VALIDATE_DEBUG_EVENT_BRANCH_RETURN,
+ /* Pruning branch as dynamically unreachable. */
+ RTE_BPF_VALIDATE_DEBUG_EVENT_BRANCH_UNREACHABLE,
/* Number of valid event values. */
RTE_BPF_VALIDATE_DEBUG_EVENT_END,
};
--
2.43.0
^ permalink raw reply related
* [PATCH v3 16/25] bpf/validate: fix BPF_JMP source range calculation
From: Marat Khalili @ 2026-06-12 10:47 UTC (permalink / raw)
To: Konstantin Ananyev; +Cc: dev, stable, Claudia Cauli
In-Reply-To: <20260612104743.6465-1-marat.khalili@huawei.com>
All two-register ordering comparison functions (`eval_jgt_jle`,
`eval_jlt_jge`, `eval_jsgt_jsle`, `eval_jslt_jsge`) were updating only
the destination register value set but not the source register one. For
instance, instruction `jgt r2, r3` should be exactly equivalent to `jlt
r3, r2`, but previously the former only updated the possible values of
r2 while the latter only updated possible values of r3. Thus the
estimate for source register was conservative and could cause false
positives.
E.g. consider the following program with the current validation code:
Tested program:
0: mov r0, #0x0
1: mov r2, #0x28
2: ldxdw r3, [r1 + 0]
3: jlt r3, #0x14, L11
4: jgt r3, #0x3c, L11
5: jslt r3, #0x14, L11
6: jsgt r3, #0x3c, L11
7: jgt r2, r3, L10 ; tested instruction
8: mov r0, #0x1
9: exit
10: mov r0, #0x2
11: exit
Pre-state:
r2: 40
r3: 20..60
...
Jump-state:
r2: 40
r3: 20..60
If tested instruction jumped from step 7 to step 10 validator expects r3
to contain values up to 60, for example 55, however for this value jump
condition r2 > r3 will never be satisfied since r2 is known to equal 40,
and thus execution would always continue to step 8 instead of jumping.
Add missing source register values update.
Introduce test harness for verifying all equivalent variations of a
comparison instruction. Add tests for all cases where both code branches
are reachable (unreachable branches will be covered by subsequent
commits).
Fixes: 8021917293d0 ("bpf: add extra validation for input BPF program")
Cc: stable@dpdk.org
Reported-by: Claudia Cauli <claudiacauli@gmail.com>
Signed-off-by: Marat Khalili <marat.khalili@huawei.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@huawei.com>
---
Depends-on: series-38434 ("bpf: introduce extensible load API")
app/test/test_bpf_validate.c | 394 +++++++++++++++++++++++++++++++----
lib/bpf/bpf_validate.c | 8 +
2 files changed, 358 insertions(+), 44 deletions(-)
diff --git a/app/test/test_bpf_validate.c b/app/test/test_bpf_validate.c
index acc238f7d324..15ccf4f6a573 100644
--- a/app/test/test_bpf_validate.c
+++ b/app/test/test_bpf_validate.c
@@ -32,6 +32,31 @@ RTE_LOG_REGISTER(test_bpf_validate_logtype, test.bpf_validate, NOTICE);
#define REGISTER_FORMAT_BUFFER_SIZE 256
#define DISASSEMBLY_FORMAT_BUFFER_SIZE 64
+#define COMPARISON_INDEX_IMMEDIATE RTE_BIT32(0)
+#define COMPARISON_INDEX_GREATER RTE_BIT32(1)
+#define COMPARISON_INDEX_INCLUSIVE RTE_BIT32(2)
+#define COMPARISON_INDEX_SIGNED RTE_BIT32(3)
+
+/* List comparison opcodes to make their index bits match constants above. */
+static const uint8_t comparisons_opcode[] = {
+ (BPF_JMP | EBPF_JLT | BPF_X),
+ (BPF_JMP | EBPF_JLT | BPF_K),
+ (BPF_JMP | BPF_JGT | BPF_X),
+ (BPF_JMP | BPF_JGT | BPF_K),
+ (BPF_JMP | EBPF_JLE | BPF_X),
+ (BPF_JMP | EBPF_JLE | BPF_K),
+ (BPF_JMP | BPF_JGE | BPF_X),
+ (BPF_JMP | BPF_JGE | BPF_K),
+ (BPF_JMP | EBPF_JSLT | BPF_X),
+ (BPF_JMP | EBPF_JSLT | BPF_K),
+ (BPF_JMP | EBPF_JSGT | BPF_X),
+ (BPF_JMP | EBPF_JSGT | BPF_K),
+ (BPF_JMP | EBPF_JSLE | BPF_X),
+ (BPF_JMP | EBPF_JSLE | BPF_K),
+ (BPF_JMP | EBPF_JSGE | BPF_X),
+ (BPF_JMP | EBPF_JSGE | BPF_K),
+};
+
/* Interval bounded by two signed values, inclusive; min <= max. */
struct signed_interval {
int64_t min;
@@ -1044,6 +1069,206 @@ verify_instruction(struct verify_instruction_param prm)
return rc;
}
+static int
+opcode_comparison_index(uint8_t opcode)
+{
+ for (int index = 0; index != RTE_DIM(comparisons_opcode); ++index)
+ if (comparisons_opcode[index] == opcode)
+ return index;
+ TEST_LOG_LINE(ERR, "Unsupported or not a comparison opcode: %hhx", opcode);
+ RTE_VERIFY(false);
+}
+
+/* Change two-register comparison verification to immediate one. */
+static bool
+make_comparison_immediate(struct verify_instruction_param *prm)
+{
+ int comparison_index = opcode_comparison_index(prm->tested_instruction.code);
+ const int64_t value = prm->pre.src.s.min;
+
+ if ((comparison_index & COMPARISON_INDEX_IMMEDIATE) != 0) {
+ TEST_LOG_LINE(ERR, "Comparison %hhx is already immediate.",
+ prm->tested_instruction.code);
+ RTE_VERIFY(false);
+ }
+
+ if (!domain_is_singleton(&prm->pre.src) || !domain_is_singleton(&prm->post.src) ||
+ !domain_is_singleton(&prm->jump.src)) {
+ TEST_LOG_LINE(DEBUG, "Cannot make immediate out of a non-singleton domain.");
+ return false;
+ }
+ if (prm->pre.src.is_pointer || prm->post.src.is_pointer || prm->jump.src.is_pointer) {
+ TEST_LOG_LINE(DEBUG, "Cannot make immediate out of a pointer.");
+ return false;
+ }
+ if (prm->post.src.s.min != value || prm->jump.src.s.min != value) {
+ TEST_LOG_LINE(DEBUG, "Cannot make immediate if the value changes.");
+ return false;
+ }
+ if (!fits_in_imm32(value)) {
+ TEST_LOG_LINE(ERR, "Cannot make immediate unless value fits in int32.");
+ return false;
+ }
+
+ comparison_index |= COMPARISON_INDEX_IMMEDIATE;
+ prm->tested_instruction.code = comparisons_opcode[comparison_index];
+ prm->tested_instruction.imm = value;
+
+ RTE_VERIFY(prm->pre.src.is_defined);
+ prm->pre.src.is_defined = false;
+
+ if (!prm->post.is_unreachable) {
+ RTE_VERIFY(prm->post.src.is_defined);
+ prm->post.src.is_defined = false;
+ }
+
+ if (!prm->jump.is_unreachable) {
+ RTE_VERIFY(prm->jump.src.is_defined);
+ prm->jump.src.is_defined = false;
+ }
+
+ return true;
+}
+
+/* Change immediate comparison verification to two-register one. */
+static void
+make_comparison_two_register(struct verify_instruction_param *prm)
+{
+ int comparison_index = opcode_comparison_index(prm->tested_instruction.code);
+ const int64_t value = prm->tested_instruction.imm;
+
+ if ((comparison_index & COMPARISON_INDEX_IMMEDIATE) == 0) {
+ TEST_LOG_LINE(ERR, "Comparison %hhx is already two-register.",
+ prm->tested_instruction.code);
+ RTE_VERIFY(false);
+ }
+
+ comparison_index &= ~COMPARISON_INDEX_IMMEDIATE;
+ prm->tested_instruction.code = comparisons_opcode[comparison_index];
+ prm->tested_instruction.imm = 0;
+
+ RTE_VERIFY(!prm->pre.src.is_defined);
+ prm->pre.src = make_singleton_domain(value);
+
+ if (!prm->post.is_unreachable) {
+ RTE_VERIFY(!prm->post.src.is_defined);
+ prm->post.src = prm->pre.src;
+ }
+
+ if (!prm->jump.is_unreachable) {
+ RTE_VERIFY(!prm->jump.src.is_defined);
+ prm->jump.src = prm->pre.src;
+ }
+}
+
+/* Change comparison verification to complement (negated result) one. */
+static void
+make_comparison_complement(struct verify_instruction_param *prm)
+{
+ int comparison_index = opcode_comparison_index(prm->tested_instruction.code);
+ comparison_index ^= COMPARISON_INDEX_GREATER | COMPARISON_INDEX_INCLUSIVE;
+ prm->tested_instruction.code = comparisons_opcode[comparison_index];
+ RTE_SWAP(prm->post, prm->jump);
+}
+
+/* Change comparison verification to converse (swapped operands) one. */
+static void
+make_comparison_converse(struct verify_instruction_param *prm)
+{
+ int comparison_index = opcode_comparison_index(prm->tested_instruction.code);
+ comparison_index ^= COMPARISON_INDEX_GREATER;
+ prm->tested_instruction.code = comparisons_opcode[comparison_index];
+ RTE_SWAP(prm->pre.dst, prm->pre.src);
+ RTE_SWAP(prm->post.dst, prm->post.src);
+ RTE_SWAP(prm->jump.dst, prm->jump.src);
+}
+
+/* Change signed comparison verification to unsigned one. */
+static void
+make_comparison_signed(struct verify_instruction_param *prm)
+{
+ int comparison_index = opcode_comparison_index(prm->tested_instruction.code);
+ if ((comparison_index & COMPARISON_INDEX_SIGNED) != 0) {
+ TEST_LOG_LINE(ERR, "Comparison %hhx is already signed.",
+ prm->tested_instruction.code);
+ RTE_VERIFY(false);
+ }
+ comparison_index |= COMPARISON_INDEX_SIGNED;
+ prm->tested_instruction.code = comparisons_opcode[comparison_index];
+}
+
+/* Verify specified two-register comparison and, if possible, immediate one. */
+static int
+verify_comparison_subcase(struct verify_instruction_param prm)
+{
+ TEST_ASSERT_SUCCESS(verify_instruction(prm), "two-register version check");
+
+ if (make_comparison_immediate(&prm))
+ TEST_ASSERT_SUCCESS(verify_instruction(prm), "immediate version check");
+
+ return TEST_SUCCESS;
+}
+
+/*
+ * Verify comparison instruction validation behaviour.
+ *
+ * Call `verify_instruction` for all valid variations of the instruction.
+ *
+ * For instance, `jgt r2, r3` verifies:
+ * * `jgt r2, r3`;
+ * * `jlt r3, r2` src and dst swapped with each other;
+ * * `jle r2, r3` with post and jump domains swapped with each other;
+ * * `jge r3, r2` with all corresponding swaps;
+ * * immediate versions of everything above where possible,
+ * that is, register on the right is an int32 scalar singleton;
+ * * signed versions of everything above if `also_signed` is true;
+ *
+ * Regardless if passed instruction compares with immediate or singleton src
+ * both cases are generated and tested.
+ */
+static int
+verify_comparison(struct verify_instruction_param prm, bool also_signed)
+{
+ fill_verify_instruction_defaults(&prm);
+
+ if (!prm.pre.src.is_defined)
+ /* Convert from immediate form to simplify further logic. */
+ make_comparison_two_register(&prm);
+
+ /* All reachable domains must be defined by this point. */
+ RTE_VERIFY(prm.pre.dst.is_defined);
+ RTE_VERIFY(prm.pre.src.is_defined);
+ if (!prm.post.is_unreachable) {
+ RTE_VERIFY(prm.post.dst.is_defined);
+ RTE_VERIFY(prm.post.src.is_defined);
+ }
+ if (!prm.jump.is_unreachable) {
+ RTE_VERIFY(prm.jump.dst.is_defined);
+ RTE_VERIFY(prm.jump.src.is_defined);
+ }
+
+ for (int make_signed = 0; make_signed <= also_signed; ++make_signed) {
+ if (make_signed)
+ make_comparison_signed(&prm);
+
+ for (int complement = false; complement <= true; ++complement) {
+
+ for (int converse = false; converse <= true; ++converse) {
+
+ TEST_ASSERT_SUCCESS(verify_comparison_subcase(prm),
+ "make_signed=%d, complement=%d, converse=%d",
+ make_signed, complement, converse);
+
+ make_comparison_converse(&prm);
+ }
+
+ make_comparison_complement(&prm);
+ }
+ }
+
+ return TEST_SUCCESS;
+}
+
/* TESTS FOR SPECIFIC INSTRUCTIONS */
@@ -1485,31 +1710,69 @@ test_jmp64_jslt_x(void)
REGISTER_FAST_TEST(bpf_validate_jmp64_jslt_x_autotest, NOHUGE_OK, ASAN_OK,
test_jmp64_jslt_x);
-/* Jump on ordering relationship with narrower range. */
+/* Jump on ordering comparisons between two ranges. */
static int
-test_jmp64_jxx_x_ordering_narrower(void)
+test_jmp64_ordering_ranges(void)
{
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ /* All ranges used are valid for both signed and unsigned comparisons. */
+ const bool also_signed = true;
+
+ /*
+ * 20 ---- dst ---- 60
+ * 10 -- src -- 40
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
- .code = (BPF_JMP | BPF_JGT | BPF_X),
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
},
.pre.dst = make_signed_domain(20, 60),
- .pre.src = make_signed_domain(30, 50),
- .post.dst = make_signed_domain(20, 50),
- .jump.dst = make_signed_domain(31, 60),
- }), "(BPF_JMP | BPF_JGT | BPF_X) check");
+ .pre.src = make_signed_domain(10, 40),
+ .jump.dst = make_signed_domain(20, 39),
+ .jump.src = make_signed_domain(21, 40),
+ }, also_signed), "strict, dst range weakly greater than src range");
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
- .code = (BPF_JMP | BPF_JGE | BPF_X),
+ .code = (BPF_JMP | EBPF_JLE | BPF_X),
},
.pre.dst = make_signed_domain(20, 60),
- .pre.src = make_signed_domain(30, 50),
- .post.dst = make_signed_domain(20, 49),
- .jump.dst = make_signed_domain(30, 60),
- }), "(BPF_JMP | BPF_JGE | BPF_X) check");
+ .pre.src = make_signed_domain(10, 40),
+ .jump.dst = make_signed_domain(20, 40),
+ .jump.src = make_signed_domain(20, 40),
+ }, also_signed), "non-strict, dst range weakly greater than src range");
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ /*
+ * 20 ---- dst ---- 60
+ * 10 -------- src -------- 70
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .pre.src = make_signed_domain(10, 70),
+ .post.src = make_signed_domain(10, 60),
+ .jump.src = make_signed_domain(21, 70),
+ }, also_signed), "strict, dst range included in src range");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_X),
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .pre.src = make_signed_domain(10, 70),
+ .post.src = make_signed_domain(10, 59),
+ .jump.src = make_signed_domain(20, 70),
+ }, also_signed), "non-strict, dst range included in src range");
+
+ /*
+ * 20 ---- dst ---- 60
+ * 30 - src - 50
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
.code = (BPF_JMP | EBPF_JLT | BPF_X),
},
@@ -1517,9 +1780,9 @@ test_jmp64_jxx_x_ordering_narrower(void)
.pre.src = make_signed_domain(30, 50),
.post.dst = make_signed_domain(30, 60),
.jump.dst = make_signed_domain(20, 49),
- }), "(BPF_JMP | EBPF_JLT | BPF_X) check");
+ }, also_signed), "strict, dst range includes src range");
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
.code = (BPF_JMP | EBPF_JLE | BPF_X),
},
@@ -1527,53 +1790,96 @@ test_jmp64_jxx_x_ordering_narrower(void)
.pre.src = make_signed_domain(30, 50),
.post.dst = make_signed_domain(31, 60),
.jump.dst = make_signed_domain(20, 50),
- }), "(BPF_JMP | EBPF_JLE | BPF_X) check");
+ }, also_signed), "non-strict, dst range includes src range");
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ /*
+ * 20 ---- dst ---- 60
+ * 40 -- src -- 70
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
- .code = (BPF_JMP | EBPF_JSGT | BPF_X),
+ .code = (BPF_JMP | EBPF_JLT | BPF_X),
},
.pre.dst = make_signed_domain(20, 60),
- .pre.src = make_signed_domain(30, 50),
- .post.dst = make_signed_domain(20, 50),
- .jump.dst = make_signed_domain(31, 60),
- }), "(BPF_JMP | EBPF_JSGT | BPF_X) check");
+ .pre.src = make_signed_domain(40, 70),
+ .post.dst = make_signed_domain(40, 60),
+ .post.src = make_signed_domain(40, 60),
+ }, also_signed), "strict, dst range weakly less than src range");
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
- .code = (BPF_JMP | EBPF_JSGE | BPF_X),
+ .code = (BPF_JMP | EBPF_JLE | BPF_X),
},
.pre.dst = make_signed_domain(20, 60),
- .pre.src = make_signed_domain(30, 50),
- .post.dst = make_signed_domain(20, 49),
- .jump.dst = make_signed_domain(30, 60),
- }), "(BPF_JMP | EBPF_JSGE | BPF_X) check");
+ .pre.src = make_signed_domain(40, 70),
+ .post.dst = make_signed_domain(41, 60),
+ .post.src = make_signed_domain(40, 59),
+ }, also_signed), "non-strict, dst range weakly less than src range");
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ return TEST_SUCCESS;
+}
+
+REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_ranges_autotest, NOHUGE_OK, ASAN_OK,
+ test_jmp64_ordering_ranges);
+
+/* Jump on ordering comparisons with singleton. */
+static int
+test_jmp64_ordering_singleton(void)
+{
+ /* All ranges used are valid for both signed and unsigned comparisons. */
+ const bool also_signed = true;
+
+ /*
+ * 20 ---- dst ---- 60
+ * imm
+ */
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
- .code = (BPF_JMP | EBPF_JSLT | BPF_X),
+ .code = (BPF_JMP | EBPF_JLT | BPF_K),
+ .imm = 40,
},
.pre.dst = make_signed_domain(20, 60),
- .pre.src = make_signed_domain(30, 50),
- .post.dst = make_signed_domain(30, 60),
- .jump.dst = make_signed_domain(20, 49),
- }), "(BPF_JMP | EBPF_JSLT | BPF_X) check");
+ .post.dst = make_signed_domain(40, 60),
+ .jump.dst = make_signed_domain(20, 39),
+ }, also_signed), "(BPF_JMP | EBPF_JLT | BPF_K) check");
- TEST_ASSERT_SUCCESS(verify_instruction((struct verify_instruction_param){
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
.tested_instruction = {
- .code = (BPF_JMP | EBPF_JSLE | BPF_X),
+ .code = (BPF_JMP | BPF_JGT | BPF_K),
+ .imm = 40,
},
.pre.dst = make_signed_domain(20, 60),
- .pre.src = make_signed_domain(30, 50),
- .post.dst = make_signed_domain(31, 60),
- .jump.dst = make_signed_domain(20, 50),
- }), "(BPF_JMP | EBPF_JSLE | BPF_X) check");
+ .post.dst = make_signed_domain(20, 40),
+ .jump.dst = make_signed_domain(41, 60),
+ }, also_signed), "(BPF_JMP | EBPF_JGT | BPF_K) check");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | EBPF_JLE | BPF_K),
+ .imm = 40,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .post.dst = make_signed_domain(41, 60),
+ .jump.dst = make_signed_domain(20, 40),
+ }, also_signed), "(BPF_JMP | EBPF_JLE | BPF_K) check");
+
+ TEST_ASSERT_SUCCESS(verify_comparison((struct verify_instruction_param){
+ .tested_instruction = {
+ .code = (BPF_JMP | BPF_JGE | BPF_K),
+ .imm = 40,
+ },
+ .pre.dst = make_signed_domain(20, 60),
+ .post.dst = make_signed_domain(20, 39),
+ .jump.dst = make_signed_domain(40, 60),
+ }, also_signed), "(BPF_JMP | EBPF_JGE | BPF_K) check");
return TEST_SUCCESS;
}
-REGISTER_FAST_TEST(bpf_validate_jmp64_jxx_x_ordering_narrower_autotest, NOHUGE_OK, ASAN_OK,
- test_jmp64_jxx_x_ordering_narrower);
+REGISTER_FAST_TEST(bpf_validate_jmp64_ordering_singleton_autotest, NOHUGE_OK, ASAN_OK,
+ test_jmp64_ordering_singleton);
/* 64-bit load from heap (should be set to unknown). */
static int
diff --git a/lib/bpf/bpf_validate.c b/lib/bpf/bpf_validate.c
index 0578029dccbb..2e535069fe4d 100644
--- a/lib/bpf/bpf_validate.c
+++ b/lib/bpf/bpf_validate.c
@@ -1522,7 +1522,9 @@ eval_jgt_jle(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
frd->u.max = RTE_MIN(frd->u.max, frs->u.max);
+ frs->u.min = RTE_MAX(frs->u.min, frd->u.min);
trd->u.min = RTE_MAX(trd->u.min, trs->u.min + 1);
+ trs->u.max = RTE_MIN(trs->u.max, trd->u.max - 1);
}
static void
@@ -1530,7 +1532,9 @@ eval_jlt_jge(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
frd->u.min = RTE_MAX(frd->u.min, frs->u.min);
+ frs->u.max = RTE_MIN(frs->u.max, frd->u.max);
trd->u.max = RTE_MIN(trd->u.max, trs->u.max - 1);
+ trs->u.min = RTE_MAX(trs->u.min, trd->u.min + 1);
}
static void
@@ -1538,7 +1542,9 @@ eval_jsgt_jsle(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
frd->s.max = RTE_MIN(frd->s.max, frs->s.max);
+ frs->s.min = RTE_MAX(frs->s.min, frd->s.min);
trd->s.min = RTE_MAX(trd->s.min, trs->s.min + 1);
+ trs->s.max = RTE_MIN(trs->s.max, trd->s.max - 1);
}
static void
@@ -1546,7 +1552,9 @@ eval_jslt_jsge(struct bpf_reg_val *trd, struct bpf_reg_val *trs,
struct bpf_reg_val *frd, struct bpf_reg_val *frs)
{
frd->s.min = RTE_MAX(frd->s.min, frs->s.min);
+ frs->s.max = RTE_MIN(frs->s.max, frd->s.max);
trd->s.max = RTE_MIN(trd->s.max, trs->s.max - 1);
+ trs->s.min = RTE_MAX(trs->s.min, trd->s.min + 1);
}
static const char *
--
2.43.0
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox