From: Luo Jie <quic_luoj@quicinc.com>
To: Andrew Lunn <andrew+netdev@lunn.ch>,
"David S. Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
Rob Herring <robh@kernel.org>,
Krzysztof Kozlowski <krzk+dt@kernel.org>,
Conor Dooley <conor+dt@kernel.org>,
Lei Wei <quic_leiwei@quicinc.com>,
Suruchi Agarwal <quic_suruchia@quicinc.com>,
Pavithra R <quic_pavir@quicinc.com>,
"Simon Horman" <horms@kernel.org>,
Jonathan Corbet <corbet@lwn.net>, Kees Cook <kees@kernel.org>,
"Gustavo A. R. Silva" <gustavoars@kernel.org>,
"Philipp Zabel" <p.zabel@pengutronix.de>
Cc: <linux-arm-msm@vger.kernel.org>, <netdev@vger.kernel.org>,
<devicetree@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
<linux-doc@vger.kernel.org>, <linux-hardening@vger.kernel.org>,
<quic_kkumarcs@quicinc.com>, <quic_linchen@quicinc.com>,
Luo Jie <quic_luoj@quicinc.com>
Subject: [PATCH net-next v8 05/14] net: ethernet: qualcomm: Initialize PPE queue management for IPQ9574
Date: Mon, 18 Aug 2025 21:14:29 +0800 [thread overview]
Message-ID: <20250818-qcom_ipq_ppe-v8-5-1d4ff641fce9@quicinc.com> (raw)
In-Reply-To: <20250818-qcom_ipq_ppe-v8-0-1d4ff641fce9@quicinc.com>
QM (queue management) configurations decide the length of PPE queues
and the queue depth for these queues which are used to drop packets
in events of congestion.
There are two types of PPE queues - unicast queues (0-255) and multicast
queues (256-299). These queue types are used to forward different types
of traffic, and are configured with different lengths.
Signed-off-by: Luo Jie <quic_luoj@quicinc.com>
---
drivers/net/ethernet/qualcomm/ppe/ppe_config.c | 184 ++++++++++++++++++++++++-
drivers/net/ethernet/qualcomm/ppe/ppe_regs.h | 85 ++++++++++++
2 files changed, 268 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/qualcomm/ppe/ppe_config.c b/drivers/net/ethernet/qualcomm/ppe/ppe_config.c
index 45b031d4dd46..53887069b432 100644
--- a/drivers/net/ethernet/qualcomm/ppe/ppe_config.c
+++ b/drivers/net/ethernet/qualcomm/ppe/ppe_config.c
@@ -43,6 +43,29 @@ struct ppe_bm_port_config {
bool dynamic;
};
+/**
+ * struct ppe_qm_queue_config - PPE queue config.
+ * @queue_start: PPE start of queue ID.
+ * @queue_end: PPE end of queue ID.
+ * @prealloc_buf: Queue dedicated buffer number.
+ * @ceil: Ceil to start drop packet from queue.
+ * @weight: Weight value.
+ * @resume_offset: Resume offset from the threshold.
+ * @dynamic: Threshold value is decided dynamically or statically.
+ *
+ * Queue configuration decides the threshold to drop packet from PPE
+ * hardware queue.
+ */
+struct ppe_qm_queue_config {
+ unsigned int queue_start;
+ unsigned int queue_end;
+ unsigned int prealloc_buf;
+ unsigned int ceil;
+ unsigned int weight;
+ unsigned int resume_offset;
+ bool dynamic;
+};
+
/* There are total 2048 buffers available in PPE, out of which some
* buffers are reserved for some specific purposes per PPE port. The
* rest of the pool of 1550 buffers are assigned to the general 'group0'
@@ -106,6 +129,40 @@ static const struct ppe_bm_port_config ipq9574_ppe_bm_port_config[] = {
},
};
+/* QM fetches the packet from PPE buffer management for transmitting the
+ * packet out. The QM group configuration limits the total number of buffers
+ * enqueued by all PPE hardware queues.
+ * There are total 2048 buffers available, out of which some buffers are
+ * dedicated to hardware exception handlers. The remaining buffers are
+ * assigned to the general 'group0', which is the group assigned to all
+ * queues by default.
+ */
+static const int ipq9574_ppe_qm_group_config = 2000;
+
+/* Default QM settings for unicast and multicast queues for IPQ9754. */
+static const struct ppe_qm_queue_config ipq9574_ppe_qm_queue_config[] = {
+ {
+ /* QM settings for unicast queues 0 to 255. */
+ .queue_start = 0,
+ .queue_end = 255,
+ .prealloc_buf = 0,
+ .ceil = 1200,
+ .weight = 7,
+ .resume_offset = 36,
+ .dynamic = true,
+ },
+ {
+ /* QM settings for multicast queues 256 to 299. */
+ .queue_start = 256,
+ .queue_end = 299,
+ .prealloc_buf = 0,
+ .ceil = 250,
+ .weight = 0,
+ .resume_offset = 36,
+ .dynamic = false,
+ },
+};
+
static int ppe_config_bm_threshold(struct ppe_device *ppe_dev, int bm_port_id,
const struct ppe_bm_port_config port_cfg)
{
@@ -193,7 +250,132 @@ static int ppe_config_bm(struct ppe_device *ppe_dev)
return ret;
}
+/* Configure PPE hardware queue depth, which is decided by the threshold
+ * of queue.
+ */
+static int ppe_config_qm(struct ppe_device *ppe_dev)
+{
+ const struct ppe_qm_queue_config *queue_cfg;
+ int ret, i, queue_id, queue_cfg_count;
+ u32 reg, multicast_queue_cfg[5];
+ u32 unicast_queue_cfg[4];
+ u32 group_cfg[3];
+
+ /* Assign the buffer number to the group 0 by default. */
+ reg = PPE_AC_GRP_CFG_TBL_ADDR;
+ ret = regmap_bulk_read(ppe_dev->regmap, reg,
+ group_cfg, ARRAY_SIZE(group_cfg));
+ if (ret)
+ goto qm_config_fail;
+
+ PPE_AC_GRP_SET_BUF_LIMIT(group_cfg, ipq9574_ppe_qm_group_config);
+
+ ret = regmap_bulk_write(ppe_dev->regmap, reg,
+ group_cfg, ARRAY_SIZE(group_cfg));
+ if (ret)
+ goto qm_config_fail;
+
+ queue_cfg = ipq9574_ppe_qm_queue_config;
+ queue_cfg_count = ARRAY_SIZE(ipq9574_ppe_qm_queue_config);
+ for (i = 0; i < queue_cfg_count; i++) {
+ queue_id = queue_cfg[i].queue_start;
+
+ /* Configure threshold for dropping packets separately for
+ * unicast and multicast PPE queues.
+ */
+ while (queue_id <= queue_cfg[i].queue_end) {
+ if (queue_id < PPE_AC_UNICAST_QUEUE_CFG_TBL_ENTRIES) {
+ reg = PPE_AC_UNICAST_QUEUE_CFG_TBL_ADDR +
+ PPE_AC_UNICAST_QUEUE_CFG_TBL_INC * queue_id;
+
+ ret = regmap_bulk_read(ppe_dev->regmap, reg,
+ unicast_queue_cfg,
+ ARRAY_SIZE(unicast_queue_cfg));
+ if (ret)
+ goto qm_config_fail;
+
+ PPE_AC_UNICAST_QUEUE_SET_EN(unicast_queue_cfg, true);
+ PPE_AC_UNICAST_QUEUE_SET_GRP_ID(unicast_queue_cfg, 0);
+ PPE_AC_UNICAST_QUEUE_SET_PRE_LIMIT(unicast_queue_cfg,
+ queue_cfg[i].prealloc_buf);
+ PPE_AC_UNICAST_QUEUE_SET_DYNAMIC(unicast_queue_cfg,
+ queue_cfg[i].dynamic);
+ PPE_AC_UNICAST_QUEUE_SET_WEIGHT(unicast_queue_cfg,
+ queue_cfg[i].weight);
+ PPE_AC_UNICAST_QUEUE_SET_THRESHOLD(unicast_queue_cfg,
+ queue_cfg[i].ceil);
+ PPE_AC_UNICAST_QUEUE_SET_GRN_RESUME(unicast_queue_cfg,
+ queue_cfg[i].resume_offset);
+
+ ret = regmap_bulk_write(ppe_dev->regmap, reg,
+ unicast_queue_cfg,
+ ARRAY_SIZE(unicast_queue_cfg));
+ if (ret)
+ goto qm_config_fail;
+ } else {
+ reg = PPE_AC_MULTICAST_QUEUE_CFG_TBL_ADDR +
+ PPE_AC_MULTICAST_QUEUE_CFG_TBL_INC * queue_id;
+
+ ret = regmap_bulk_read(ppe_dev->regmap, reg,
+ multicast_queue_cfg,
+ ARRAY_SIZE(multicast_queue_cfg));
+ if (ret)
+ goto qm_config_fail;
+
+ PPE_AC_MULTICAST_QUEUE_SET_EN(multicast_queue_cfg, true);
+ PPE_AC_MULTICAST_QUEUE_SET_GRN_GRP_ID(multicast_queue_cfg, 0);
+ PPE_AC_MULTICAST_QUEUE_SET_GRN_PRE_LIMIT(multicast_queue_cfg,
+ queue_cfg[i].prealloc_buf);
+ PPE_AC_MULTICAST_QUEUE_SET_GRN_THRESHOLD(multicast_queue_cfg,
+ queue_cfg[i].ceil);
+ PPE_AC_MULTICAST_QUEUE_SET_GRN_RESUME(multicast_queue_cfg,
+ queue_cfg[i].resume_offset);
+
+ ret = regmap_bulk_write(ppe_dev->regmap, reg,
+ multicast_queue_cfg,
+ ARRAY_SIZE(multicast_queue_cfg));
+ if (ret)
+ goto qm_config_fail;
+ }
+
+ /* Enable enqueue. */
+ reg = PPE_ENQ_OPR_TBL_ADDR + PPE_ENQ_OPR_TBL_INC * queue_id;
+ ret = regmap_clear_bits(ppe_dev->regmap, reg,
+ PPE_ENQ_OPR_TBL_ENQ_DISABLE);
+ if (ret)
+ goto qm_config_fail;
+
+ /* Enable dequeue. */
+ reg = PPE_DEQ_OPR_TBL_ADDR + PPE_DEQ_OPR_TBL_INC * queue_id;
+ ret = regmap_clear_bits(ppe_dev->regmap, reg,
+ PPE_DEQ_OPR_TBL_DEQ_DISABLE);
+ if (ret)
+ goto qm_config_fail;
+
+ queue_id++;
+ }
+ }
+
+ /* Enable queue counter for all PPE hardware queues. */
+ ret = regmap_set_bits(ppe_dev->regmap, PPE_EG_BRIDGE_CONFIG_ADDR,
+ PPE_EG_BRIDGE_CONFIG_QUEUE_CNT_EN);
+ if (ret)
+ goto qm_config_fail;
+
+ return 0;
+
+qm_config_fail:
+ dev_err(ppe_dev->dev, "PPE QM config error %d\n", ret);
+ return ret;
+}
+
int ppe_hw_config(struct ppe_device *ppe_dev)
{
- return ppe_config_bm(ppe_dev);
+ int ret;
+
+ ret = ppe_config_bm(ppe_dev);
+ if (ret)
+ return ret;
+
+ return ppe_config_qm(ppe_dev);
}
diff --git a/drivers/net/ethernet/qualcomm/ppe/ppe_regs.h b/drivers/net/ethernet/qualcomm/ppe/ppe_regs.h
index b89d717fdae8..ca256fe2a321 100644
--- a/drivers/net/ethernet/qualcomm/ppe/ppe_regs.h
+++ b/drivers/net/ethernet/qualcomm/ppe/ppe_regs.h
@@ -9,6 +9,16 @@
#include <linux/bitfield.h>
+/* PPE queue counters enable/disable control. */
+#define PPE_EG_BRIDGE_CONFIG_ADDR 0x20044
+#define PPE_EG_BRIDGE_CONFIG_QUEUE_CNT_EN BIT(2)
+
+/* Table addresses for per-queue dequeue setting. */
+#define PPE_DEQ_OPR_TBL_ADDR 0x430000
+#define PPE_DEQ_OPR_TBL_ENTRIES 300
+#define PPE_DEQ_OPR_TBL_INC 0x10
+#define PPE_DEQ_OPR_TBL_DEQ_DISABLE BIT(0)
+
/* There are 15 BM ports and 4 BM groups supported by PPE.
* BM port (0-7) is for EDMA port 0, BM port (8-13) is for
* PPE physical port 1-6 and BM port 14 is for EIP port.
@@ -56,4 +66,79 @@
FIELD_MODIFY(PPE_BM_PORT_FC_W1_DYNAMIC, (tbl_cfg) + 0x1, value)
#define PPE_BM_PORT_FC_SET_PRE_ALLOC(tbl_cfg, value) \
FIELD_MODIFY(PPE_BM_PORT_FC_W1_PRE_ALLOC, (tbl_cfg) + 0x1, value)
+
+/* PPE unicast queue (0-255) configurations. */
+#define PPE_AC_UNICAST_QUEUE_CFG_TBL_ADDR 0x848000
+#define PPE_AC_UNICAST_QUEUE_CFG_TBL_ENTRIES 256
+#define PPE_AC_UNICAST_QUEUE_CFG_TBL_INC 0x10
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_EN BIT(0)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_WRED_EN BIT(1)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_FC_EN BIT(2)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_CLR_AWARE BIT(3)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_GRP_ID GENMASK(5, 4)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_PRE_LIMIT GENMASK(16, 6)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_DYNAMIC BIT(17)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_WEIGHT GENMASK(20, 18)
+#define PPE_AC_UNICAST_QUEUE_CFG_W0_THRESHOLD GENMASK(31, 21)
+#define PPE_AC_UNICAST_QUEUE_CFG_W3_GRN_RESUME GENMASK(23, 13)
+
+#define PPE_AC_UNICAST_QUEUE_SET_EN(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_UNICAST_QUEUE_CFG_W0_EN, tbl_cfg, value)
+#define PPE_AC_UNICAST_QUEUE_SET_GRP_ID(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_UNICAST_QUEUE_CFG_W0_GRP_ID, tbl_cfg, value)
+#define PPE_AC_UNICAST_QUEUE_SET_PRE_LIMIT(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_UNICAST_QUEUE_CFG_W0_PRE_LIMIT, tbl_cfg, value)
+#define PPE_AC_UNICAST_QUEUE_SET_DYNAMIC(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_UNICAST_QUEUE_CFG_W0_DYNAMIC, tbl_cfg, value)
+#define PPE_AC_UNICAST_QUEUE_SET_WEIGHT(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_UNICAST_QUEUE_CFG_W0_WEIGHT, tbl_cfg, value)
+#define PPE_AC_UNICAST_QUEUE_SET_THRESHOLD(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_UNICAST_QUEUE_CFG_W0_THRESHOLD, tbl_cfg, value)
+#define PPE_AC_UNICAST_QUEUE_SET_GRN_RESUME(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_UNICAST_QUEUE_CFG_W3_GRN_RESUME, (tbl_cfg) + 0x3, value)
+
+/* PPE multicast queue (256-299) configurations. */
+#define PPE_AC_MULTICAST_QUEUE_CFG_TBL_ADDR 0x84a000
+#define PPE_AC_MULTICAST_QUEUE_CFG_TBL_ENTRIES 44
+#define PPE_AC_MULTICAST_QUEUE_CFG_TBL_INC 0x10
+#define PPE_AC_MULTICAST_QUEUE_CFG_W0_EN BIT(0)
+#define PPE_AC_MULTICAST_QUEUE_CFG_W0_FC_EN BIT(1)
+#define PPE_AC_MULTICAST_QUEUE_CFG_W0_CLR_AWARE BIT(2)
+#define PPE_AC_MULTICAST_QUEUE_CFG_W0_GRP_ID GENMASK(4, 3)
+#define PPE_AC_MULTICAST_QUEUE_CFG_W0_PRE_LIMIT GENMASK(15, 5)
+#define PPE_AC_MULTICAST_QUEUE_CFG_W0_THRESHOLD GENMASK(26, 16)
+#define PPE_AC_MULTICAST_QUEUE_CFG_W2_RESUME GENMASK(17, 7)
+
+#define PPE_AC_MULTICAST_QUEUE_SET_EN(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_MULTICAST_QUEUE_CFG_W0_EN, tbl_cfg, value)
+#define PPE_AC_MULTICAST_QUEUE_SET_GRN_GRP_ID(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_MULTICAST_QUEUE_CFG_W0_GRP_ID, tbl_cfg, value)
+#define PPE_AC_MULTICAST_QUEUE_SET_GRN_PRE_LIMIT(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_MULTICAST_QUEUE_CFG_W0_PRE_LIMIT, tbl_cfg, value)
+#define PPE_AC_MULTICAST_QUEUE_SET_GRN_THRESHOLD(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_MULTICAST_QUEUE_CFG_W0_THRESHOLD, tbl_cfg, value)
+#define PPE_AC_MULTICAST_QUEUE_SET_GRN_RESUME(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_MULTICAST_QUEUE_CFG_W2_RESUME, (tbl_cfg) + 0x2, value)
+
+/* PPE admission control group (0-3) configurations */
+#define PPE_AC_GRP_CFG_TBL_ADDR 0x84c000
+#define PPE_AC_GRP_CFG_TBL_ENTRIES 0x4
+#define PPE_AC_GRP_CFG_TBL_INC 0x10
+#define PPE_AC_GRP_W0_AC_EN BIT(0)
+#define PPE_AC_GRP_W0_AC_FC_EN BIT(1)
+#define PPE_AC_GRP_W0_CLR_AWARE BIT(2)
+#define PPE_AC_GRP_W0_THRESHOLD_LOW GENMASK(31, 25)
+#define PPE_AC_GRP_W1_THRESHOLD_HIGH GENMASK(3, 0)
+#define PPE_AC_GRP_W1_BUF_LIMIT GENMASK(14, 4)
+#define PPE_AC_GRP_W2_RESUME_GRN GENMASK(15, 5)
+#define PPE_AC_GRP_W2_PRE_ALLOC GENMASK(26, 16)
+
+#define PPE_AC_GRP_SET_BUF_LIMIT(tbl_cfg, value) \
+ FIELD_MODIFY(PPE_AC_GRP_W1_BUF_LIMIT, (tbl_cfg) + 0x1, value)
+
+/* Table addresses for per-queue enqueue setting. */
+#define PPE_ENQ_OPR_TBL_ADDR 0x85c000
+#define PPE_ENQ_OPR_TBL_ENTRIES 300
+#define PPE_ENQ_OPR_TBL_INC 0x10
+#define PPE_ENQ_OPR_TBL_ENQ_DISABLE BIT(0)
#endif
--
2.34.1
next prev parent reply other threads:[~2025-08-18 13:15 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-08-18 13:14 [PATCH net-next v8 00/14] Add PPE driver for Qualcomm IPQ9574 SoC Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 01/14] dt-bindings: net: Add PPE " Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 02/14] docs: networking: Add PPE driver documentation " Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 03/14] net: ethernet: qualcomm: Add PPE driver for " Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 04/14] net: ethernet: qualcomm: Initialize PPE buffer management for IPQ9574 Luo Jie
2025-08-18 13:14 ` Luo Jie [this message]
2025-08-18 13:14 ` [PATCH net-next v8 06/14] net: ethernet: qualcomm: Initialize the PPE scheduler settings Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 07/14] net: ethernet: qualcomm: Initialize PPE queue settings Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 08/14] net: ethernet: qualcomm: Initialize PPE service code settings Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 09/14] net: ethernet: qualcomm: Initialize PPE port control settings Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 10/14] net: ethernet: qualcomm: Initialize PPE RSS hash settings Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 11/14] net: ethernet: qualcomm: Initialize PPE queue to Ethernet DMA ring mapping Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 12/14] net: ethernet: qualcomm: Initialize PPE L2 bridge settings Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 13/14] net: ethernet: qualcomm: Add PPE debugfs support for PPE counters Luo Jie
2025-08-18 13:14 ` [PATCH net-next v8 14/14] MAINTAINERS: Add maintainer for Qualcomm PPE driver Luo Jie
2025-08-21 10:50 ` [PATCH net-next v8 00/14] Add PPE driver for Qualcomm IPQ9574 SoC patchwork-bot+netdevbpf
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250818-qcom_ipq_ppe-v8-5-1d4ff641fce9@quicinc.com \
--to=quic_luoj@quicinc.com \
--cc=andrew+netdev@lunn.ch \
--cc=conor+dt@kernel.org \
--cc=corbet@lwn.net \
--cc=davem@davemloft.net \
--cc=devicetree@vger.kernel.org \
--cc=edumazet@google.com \
--cc=gustavoars@kernel.org \
--cc=horms@kernel.org \
--cc=kees@kernel.org \
--cc=krzk+dt@kernel.org \
--cc=kuba@kernel.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-hardening@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=p.zabel@pengutronix.de \
--cc=pabeni@redhat.com \
--cc=quic_kkumarcs@quicinc.com \
--cc=quic_leiwei@quicinc.com \
--cc=quic_linchen@quicinc.com \
--cc=quic_pavir@quicinc.com \
--cc=quic_suruchia@quicinc.com \
--cc=robh@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).