2022-11-02 01:00:14

by Daniel Golle

[permalink] [raw]
Subject: [PATCH v3] net: ethernet: mediatek: ppe: add support for flow accounting

The PPE units found in MT7622 and newer support packet and byte
accounting of hw-offloaded flows. Add support for reading those
counters as found in MediaTek's SDK[1].

[1]: https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/+/bc6a6a375c800dc2b80e1a325a2c732d1737df92
Signed-off-by: Daniel Golle <[email protected]>
---
v3: don't bother to set 'false' values in any zero-initialized struct
use mtk_foe_entry_ib2
both changes were requested by Felix Fietkau

v2: fix wrong variable name in return value check spotted by Denis Kirjanov


drivers/net/ethernet/mediatek/mtk_eth_soc.c | 7 +-
drivers/net/ethernet/mediatek/mtk_eth_soc.h | 1 +
drivers/net/ethernet/mediatek/mtk_ppe.c | 110 +++++++++++++++++-
drivers/net/ethernet/mediatek/mtk_ppe.h | 23 +++-
.../net/ethernet/mediatek/mtk_ppe_debugfs.c | 9 +-
.../net/ethernet/mediatek/mtk_ppe_offload.c | 7 ++
drivers/net/ethernet/mediatek/mtk_ppe_regs.h | 14 +++
7 files changed, 166 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index 789268b15106ec..c059315d9c080d 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -4140,7 +4140,9 @@ static int mtk_probe(struct platform_device *pdev)
u32 ppe_addr = eth->soc->reg_map->ppe_base + i * 0x400;

eth->ppe[i] = mtk_ppe_init(eth, eth->base + ppe_addr,
- eth->soc->offload_version, i);
+ eth->soc->offload_version, i,
+ eth->soc->has_accounting);
+
if (!eth->ppe[i]) {
err = -ENOMEM;
goto err_free_dev;
@@ -4259,6 +4261,7 @@ static const struct mtk_soc_data mt7622_data = {
.required_pctl = false,
.offload_version = 2,
.hash_offset = 2,
+ .has_accounting = true,
.foe_entry_size = sizeof(struct mtk_foe_entry) - 16,
.txrx = {
.txd_size = sizeof(struct mtk_tx_dma),
@@ -4296,6 +4299,7 @@ static const struct mtk_soc_data mt7629_data = {
.hw_features = MTK_HW_FEATURES,
.required_clks = MT7629_CLKS_BITMAP,
.required_pctl = false,
+ .has_accounting = true,
.txrx = {
.txd_size = sizeof(struct mtk_tx_dma),
.rxd_size = sizeof(struct mtk_rx_dma),
@@ -4315,6 +4319,7 @@ static const struct mtk_soc_data mt7986_data = {
.required_pctl = false,
.hash_offset = 4,
.foe_entry_size = sizeof(struct mtk_foe_entry),
+ .has_accounting = true,
.txrx = {
.txd_size = sizeof(struct mtk_tx_dma_v2),
.rxd_size = sizeof(struct mtk_rx_dma_v2),
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.h b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
index 589f27ddc40163..6c2a0a1826c3c7 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.h
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.h
@@ -993,6 +993,7 @@ struct mtk_soc_data {
u8 hash_offset;
u16 foe_entry_size;
netdev_features_t hw_features;
+ bool has_accounting;
struct {
u32 txd_size;
u32 rxd_size;
diff --git a/drivers/net/ethernet/mediatek/mtk_ppe.c b/drivers/net/ethernet/mediatek/mtk_ppe.c
index 2d8ca99f2467ff..349cdfb92ae029 100644
--- a/drivers/net/ethernet/mediatek/mtk_ppe.c
+++ b/drivers/net/ethernet/mediatek/mtk_ppe.c
@@ -74,6 +74,46 @@ static int mtk_ppe_wait_busy(struct mtk_ppe *ppe)
return ret;
}

+static int mtk_ppe_mib_wait_busy(struct mtk_ppe *ppe)
+{
+ int ret;
+ u32 val;
+
+ ret = readl_poll_timeout(ppe->base + MTK_PPE_MIB_SER_CR, val,
+ !(val & MTK_PPE_MIB_SER_CR_ST),
+ 20, MTK_PPE_WAIT_TIMEOUT_US);
+
+ if (ret)
+ dev_err(ppe->dev, "MIB table busy");
+
+ return ret;
+}
+
+int mtk_mib_entry_read(struct mtk_ppe *ppe, u16 index, u64 *bytes, u64 *packets)
+{
+ u32 val, cnt_r0, cnt_r1, cnt_r2;
+ u32 byte_cnt_low, byte_cnt_high, pkt_cnt_low, pkt_cnt_high;
+
+ val = FIELD_PREP(MTK_PPE_MIB_SER_CR_ADDR, index) | MTK_PPE_MIB_SER_CR_ST;
+ ppe_w32(ppe, MTK_PPE_MIB_SER_CR, val);
+
+ if (mtk_ppe_mib_wait_busy(ppe))
+ return -ETIMEDOUT;
+
+ cnt_r0 = readl(ppe->base + MTK_PPE_MIB_SER_R0);
+ cnt_r1 = readl(ppe->base + MTK_PPE_MIB_SER_R1);
+ cnt_r2 = readl(ppe->base + MTK_PPE_MIB_SER_R2);
+
+ byte_cnt_low = FIELD_GET(MTK_PPE_MIB_SER_R0_BYTE_CNT_LOW, cnt_r0);
+ byte_cnt_high = FIELD_GET(MTK_PPE_MIB_SER_R1_BYTE_CNT_HIGH, cnt_r1);
+ pkt_cnt_low = FIELD_GET(MTK_PPE_MIB_SER_R1_PKT_CNT_LOW, cnt_r1);
+ pkt_cnt_high = FIELD_GET(MTK_PPE_MIB_SER_R2_PKT_CNT_HIGH, cnt_r2);
+ *bytes = ((u64)byte_cnt_high << 32) | byte_cnt_low;
+ *packets = (pkt_cnt_high << 16) | pkt_cnt_low;
+
+ return 0;
+}
+
static void mtk_ppe_cache_clear(struct mtk_ppe *ppe)
{
ppe_set(ppe, MTK_PPE_CACHE_CTL, MTK_PPE_CACHE_CTL_CLEAR);
@@ -438,6 +478,13 @@ __mtk_foe_entry_clear(struct mtk_ppe *ppe, struct mtk_flow_entry *entry)
hwe->ib1 &= ~MTK_FOE_IB1_STATE;
hwe->ib1 |= FIELD_PREP(MTK_FOE_IB1_STATE, MTK_FOE_STATE_INVALID);
dma_wmb();
+ if (ppe->accounting) {
+ struct mtk_foe_accounting *acct;
+
+ acct = ppe->acct_table + entry->hash * sizeof(*acct);
+ acct->packets = 0;
+ acct->bytes = 0;
+ }
}
entry->hash = 0xffff;

@@ -545,6 +592,9 @@ __mtk_foe_entry_commit(struct mtk_ppe *ppe, struct mtk_foe_entry *entry,
wmb();
hwe->ib1 = entry->ib1;

+ if (ppe->accounting)
+ *mtk_foe_entry_ib2(eth, hwe) |= MTK_FOE_IB2_MIB_CNT;
+
dma_wmb();

mtk_ppe_cache_clear(ppe);
@@ -710,14 +760,42 @@ int mtk_foe_entry_idle_time(struct mtk_ppe *ppe, struct mtk_flow_entry *entry)
return __mtk_foe_entry_idle_time(ppe, entry->data.ib1);
}

+struct mtk_foe_accounting *mtk_foe_entry_get_mib(struct mtk_ppe *ppe, u32 index,
+ struct mtk_foe_accounting *diff)
+{
+ struct mtk_foe_accounting *acct;
+ int size = sizeof(struct mtk_foe_accounting);
+ u64 bytes, packets;
+
+ if (!ppe->accounting)
+ return NULL;
+
+ if (mtk_mib_entry_read(ppe, index, &bytes, &packets))
+ return NULL;
+
+ acct = ppe->acct_table + index * size;
+
+ acct->bytes += bytes;
+ acct->packets += packets;
+
+ if (diff) {
+ diff->bytes = bytes;
+ diff->packets = packets;
+ }
+
+ return acct;
+}
+
struct mtk_ppe *mtk_ppe_init(struct mtk_eth *eth, void __iomem *base,
- int version, int index)
+ int version, int index, bool accounting)
{
const struct mtk_soc_data *soc = eth->soc;
struct device *dev = eth->dev;
struct mtk_ppe *ppe;
u32 foe_flow_size;
void *foe;
+ struct mtk_mib_entry *mib;
+ struct mtk_foe_accounting *acct;

ppe = devm_kzalloc(dev, sizeof(*ppe), GFP_KERNEL);
if (!ppe)
@@ -732,6 +810,7 @@ struct mtk_ppe *mtk_ppe_init(struct mtk_eth *eth, void __iomem *base,
ppe->eth = eth;
ppe->dev = dev;
ppe->version = version;
+ ppe->accounting = accounting;

foe = dmam_alloc_coherent(ppe->dev,
MTK_PPE_ENTRIES * soc->foe_entry_size,
@@ -747,6 +826,25 @@ struct mtk_ppe *mtk_ppe_init(struct mtk_eth *eth, void __iomem *base,
if (!ppe->foe_flow)
return NULL;

+ if (accounting) {
+ mib = dmam_alloc_coherent(ppe->dev, MTK_PPE_ENTRIES * sizeof(*mib),
+ &ppe->mib_phys, GFP_KERNEL);
+ if (!mib)
+ return NULL;
+
+ memset(mib, 0, MTK_PPE_ENTRIES * sizeof(*mib));
+
+ ppe->mib_table = mib;
+
+ acct = devm_kzalloc(dev, MTK_PPE_ENTRIES * sizeof(*acct),
+ GFP_KERNEL);
+
+ if (!acct)
+ return NULL;
+
+ ppe->acct_table = acct;
+ }
+
mtk_ppe_debugfs_init(ppe, index);

return ppe;
@@ -861,6 +959,16 @@ void mtk_ppe_start(struct mtk_ppe *ppe)
ppe_w32(ppe, MTK_PPE_DEFAULT_CPU_PORT1, 0xcb777);
ppe_w32(ppe, MTK_PPE_SBW_CTRL, 0x7f);
}
+
+ if (ppe->accounting && ppe->mib_phys) {
+ ppe_w32(ppe, MTK_PPE_MIB_TB_BASE, ppe->mib_phys);
+ ppe_m32(ppe, MTK_PPE_MIB_CFG, MTK_PPE_MIB_CFG_EN,
+ MTK_PPE_MIB_CFG_EN);
+ ppe_m32(ppe, MTK_PPE_MIB_CFG, MTK_PPE_MIB_CFG_RD_CLR,
+ MTK_PPE_MIB_CFG_RD_CLR);
+ ppe_m32(ppe, MTK_PPE_MIB_CACHE_CTL, MTK_PPE_MIB_CACHE_CTL_EN,
+ MTK_PPE_MIB_CFG_RD_CLR);
+ }
}

int mtk_ppe_stop(struct mtk_ppe *ppe)
diff --git a/drivers/net/ethernet/mediatek/mtk_ppe.h b/drivers/net/ethernet/mediatek/mtk_ppe.h
index 0b7a67a958e4ca..b77509a2b0fa30 100644
--- a/drivers/net/ethernet/mediatek/mtk_ppe.h
+++ b/drivers/net/ethernet/mediatek/mtk_ppe.h
@@ -57,6 +57,7 @@ enum {
#define MTK_FOE_IB2_MULTICAST BIT(8)

#define MTK_FOE_IB2_WDMA_QID2 GENMASK(13, 12)
+#define MTK_FOE_IB2_MIB_CNT BIT(15)
#define MTK_FOE_IB2_WDMA_DEVIDX BIT(16)
#define MTK_FOE_IB2_WDMA_WINFO BIT(17)

@@ -284,16 +285,34 @@ struct mtk_flow_entry {
unsigned long cookie;
};

+struct mtk_mib_entry {
+ u32 byt_cnt_l;
+ u16 byt_cnt_h;
+ u32 pkt_cnt_l;
+ u8 pkt_cnt_h;
+ u8 _rsv0;
+ u32 _rsv1;
+} __packed;
+
+struct mtk_foe_accounting {
+ u64 bytes;
+ u64 packets;
+};
+
struct mtk_ppe {
struct mtk_eth *eth;
struct device *dev;
void __iomem *base;
int version;
char dirname[5];
+ bool accounting;

void *foe_table;
dma_addr_t foe_phys;

+ struct mtk_mib_entry *mib_table;
+ dma_addr_t mib_phys;
+
u16 foe_check_time[MTK_PPE_ENTRIES];
struct hlist_head *foe_flow;

@@ -303,7 +322,7 @@ struct mtk_ppe {
};

struct mtk_ppe *mtk_ppe_init(struct mtk_eth *eth, void __iomem *base,
- int version, int index);
+ int version, int index, bool accounting);
void mtk_ppe_start(struct mtk_ppe *ppe);
int mtk_ppe_stop(struct mtk_ppe *ppe);

@@ -354,5 +373,7 @@ int mtk_foe_entry_commit(struct mtk_ppe *ppe, struct mtk_flow_entry *entry);
void mtk_foe_entry_clear(struct mtk_ppe *ppe, struct mtk_flow_entry *entry);
int mtk_foe_entry_idle_time(struct mtk_ppe *ppe, struct mtk_flow_entry *entry);
int mtk_ppe_debugfs_init(struct mtk_ppe *ppe, int index);
+struct mtk_foe_accounting *mtk_foe_entry_get_mib(struct mtk_ppe *ppe, u32 index,
+ struct mtk_foe_accounting *diff);

#endif
diff --git a/drivers/net/ethernet/mediatek/mtk_ppe_debugfs.c b/drivers/net/ethernet/mediatek/mtk_ppe_debugfs.c
index 391b071bcff38d..39775740340b9e 100644
--- a/drivers/net/ethernet/mediatek/mtk_ppe_debugfs.c
+++ b/drivers/net/ethernet/mediatek/mtk_ppe_debugfs.c
@@ -82,6 +82,7 @@ mtk_ppe_debugfs_foe_show(struct seq_file *m, void *private, bool bind)
struct mtk_foe_entry *entry = mtk_foe_get_entry(ppe, i);
struct mtk_foe_mac_info *l2;
struct mtk_flow_addr_info ai = {};
+ struct mtk_foe_accounting *acct;
unsigned char h_source[ETH_ALEN];
unsigned char h_dest[ETH_ALEN];
int type, state;
@@ -95,6 +96,8 @@ mtk_ppe_debugfs_foe_show(struct seq_file *m, void *private, bool bind)
if (bind && state != MTK_FOE_STATE_BIND)
continue;

+ acct = mtk_foe_entry_get_mib(ppe, i, NULL);
+
type = FIELD_GET(MTK_FOE_IB1_PACKET_TYPE, entry->ib1);
seq_printf(m, "%05x %s %7s", i,
mtk_foe_entry_state_str(state),
@@ -153,9 +156,11 @@ mtk_ppe_debugfs_foe_show(struct seq_file *m, void *private, bool bind)
*((__be16 *)&h_dest[4]) = htons(l2->dest_mac_lo);

seq_printf(m, " eth=%pM->%pM etype=%04x"
- " vlan=%d,%d ib1=%08x ib2=%08x\n",
+ " vlan=%d,%d ib1=%08x ib2=%08x"
+ " packets=%lld bytes=%lld\n",
h_source, h_dest, ntohs(l2->etype),
- l2->vlan1, l2->vlan2, entry->ib1, ib2);
+ l2->vlan1, l2->vlan2, entry->ib1, ib2,
+ acct->packets, acct->bytes);
}

return 0;
diff --git a/drivers/net/ethernet/mediatek/mtk_ppe_offload.c b/drivers/net/ethernet/mediatek/mtk_ppe_offload.c
index 28bbd1df3e3059..42b5f279cfa5bb 100644
--- a/drivers/net/ethernet/mediatek/mtk_ppe_offload.c
+++ b/drivers/net/ethernet/mediatek/mtk_ppe_offload.c
@@ -491,6 +491,7 @@ static int
mtk_flow_offload_stats(struct mtk_eth *eth, struct flow_cls_offload *f)
{
struct mtk_flow_entry *entry;
+ struct mtk_foe_accounting diff;
u32 idle;

entry = rhashtable_lookup(&eth->flow_table, &f->cookie,
@@ -501,6 +502,12 @@ mtk_flow_offload_stats(struct mtk_eth *eth, struct flow_cls_offload *f)
idle = mtk_foe_entry_idle_time(eth->ppe[entry->ppe_index], entry);
f->stats.lastused = jiffies - idle * HZ;

+ if (entry->hash != 0xFFFF) {
+ mtk_foe_entry_get_mib(eth->ppe[entry->ppe_index], entry->hash, &diff);
+ f->stats.pkts += diff.packets;
+ f->stats.bytes += diff.bytes;
+ }
+
return 0;
}

diff --git a/drivers/net/ethernet/mediatek/mtk_ppe_regs.h b/drivers/net/ethernet/mediatek/mtk_ppe_regs.h
index 59596d823d8b8a..be804571b825e7 100644
--- a/drivers/net/ethernet/mediatek/mtk_ppe_regs.h
+++ b/drivers/net/ethernet/mediatek/mtk_ppe_regs.h
@@ -143,6 +143,20 @@ enum {

#define MTK_PPE_MIB_TB_BASE 0x338

+#define MTK_PPE_MIB_SER_CR 0x33C
+#define MTK_PPE_MIB_SER_CR_ST BIT(16)
+#define MTK_PPE_MIB_SER_CR_ADDR GENMASK(13, 0)
+
+#define MTK_PPE_MIB_SER_R0 0x340
+#define MTK_PPE_MIB_SER_R0_BYTE_CNT_LOW GENMASK(31, 0)
+
+#define MTK_PPE_MIB_SER_R1 0x344
+#define MTK_PPE_MIB_SER_R1_PKT_CNT_LOW GENMASK(31, 16)
+#define MTK_PPE_MIB_SER_R1_BYTE_CNT_HIGH GENMASK(15, 0)
+
+#define MTK_PPE_MIB_SER_R2 0x348
+#define MTK_PPE_MIB_SER_R2_PKT_CNT_HIGH GENMASK(23, 0)
+
#define MTK_PPE_MIB_CACHE_CTL 0x350
#define MTK_PPE_MIB_CACHE_CTL_EN BIT(0)
#define MTK_PPE_MIB_CACHE_CTL_FLUSH BIT(2)
--
2.38.1



2022-11-02 01:23:33

by Andrew Lunn

[permalink] [raw]
Subject: Re: [PATCH v3] net: ethernet: mediatek: ppe: add support for flow accounting

On Wed, Nov 02, 2022 at 12:42:40AM +0000, Daniel Golle wrote:
> The PPE units found in MT7622 and newer support packet and byte
> accounting of hw-offloaded flows. Add support for reading those
> counters as found in MediaTek's SDK[1].
>
> [1]: https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/+/bc6a6a375c800dc2b80e1a325a2c732d1737df92
> Signed-off-by: Daniel Golle <[email protected]>

Sorry, but NACK.

You have not explained why doing this correctly via ethtool -S cannot
be done. debugfs is a vendor crap way of doing this.

Andrew

2022-11-02 01:29:25

by Daniel Golle

[permalink] [raw]
Subject: Re: [PATCH v3] net: ethernet: mediatek: ppe: add support for flow accounting

Hi Andrew,

On Wed, Nov 02, 2022 at 01:51:21AM +0100, Andrew Lunn wrote:
> On Wed, Nov 02, 2022 at 12:42:40AM +0000, Daniel Golle wrote:
> > The PPE units found in MT7622 and newer support packet and byte
> > accounting of hw-offloaded flows. Add support for reading those
> > counters as found in MediaTek's SDK[1].
> >
> > [1]: https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/+/bc6a6a375c800dc2b80e1a325a2c732d1737df92
> > Signed-off-by: Daniel Golle <[email protected]>
>
> Sorry, but NACK.
>
> You have not explained why doing this correctly via ethtool -S cannot
> be done. debugfs is a vendor crap way of doing this.

The debugfs interface is pre-existing and **in addition** to the
standard Linux interfaces which are also provided. It is true that
the debugfs interface in this case doesn't provide much additional
value apart from having the counter listed next to the hardware-
specific hashtable keys. As the debugfs interface for now aims to
be as complete as possible, naturally there is some redundance of
things which can also be accessed using other (standard) interfaces.

Anyway. See function mtk_flow_offload_stats which now also reports
counters.


Cheers


Daniel

2022-11-02 02:02:40

by Andrew Lunn

[permalink] [raw]
Subject: Re: [PATCH v3] net: ethernet: mediatek: ppe: add support for flow accounting

On Wed, Nov 02, 2022 at 01:03:08AM +0000, Daniel Golle wrote:
> Hi Andrew,
>
> On Wed, Nov 02, 2022 at 01:51:21AM +0100, Andrew Lunn wrote:
> > On Wed, Nov 02, 2022 at 12:42:40AM +0000, Daniel Golle wrote:
> > > The PPE units found in MT7622 and newer support packet and byte
> > > accounting of hw-offloaded flows. Add support for reading those
> > > counters as found in MediaTek's SDK[1].
> > >
> > > [1]: https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/+/bc6a6a375c800dc2b80e1a325a2c732d1737df92
> > > Signed-off-by: Daniel Golle <[email protected]>
> >
> > Sorry, but NACK.
> >
> > You have not explained why doing this correctly via ethtool -S cannot
> > be done. debugfs is a vendor crap way of doing this.
>
> The debugfs interface is pre-existing and **in addition** to the
> standard Linux interfaces which are also provided. It is true that
> the debugfs interface in this case doesn't provide much additional
> value apart from having the counter listed next to the hardware-
> specific hashtable keys. As the debugfs interface for now aims to
> be as complete as possible, naturally there is some redundance of
> things which can also be accessed using other (standard) interfaces.

debugfs is by definition unstable. It is not ABI. Anything using it is
expected to break in the near future when it changes its layout. It is
also totally option, you cannot expect it to be mounted.

I hope you don't have any user space code using it.

Maybe i should submit a patch which just for the fun of it rearranged
the order in debugfs and change the file name?

Andrew

2022-11-02 06:24:54

by Felix Fietkau

[permalink] [raw]
Subject: Re: [PATCH v3] net: ethernet: mediatek: ppe: add support for flow accounting

On 02.11.22 02:26, Andrew Lunn wrote:
> On Wed, Nov 02, 2022 at 01:03:08AM +0000, Daniel Golle wrote:
>> Hi Andrew,
>>
>> On Wed, Nov 02, 2022 at 01:51:21AM +0100, Andrew Lunn wrote:
>> > On Wed, Nov 02, 2022 at 12:42:40AM +0000, Daniel Golle wrote:
>> > > The PPE units found in MT7622 and newer support packet and byte
>> > > accounting of hw-offloaded flows. Add support for reading those
>> > > counters as found in MediaTek's SDK[1].
>> > >
>> > > [1]: https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/+/bc6a6a375c800dc2b80e1a325a2c732d1737df92
>> > > Signed-off-by: Daniel Golle <[email protected]>
>> >
>> > Sorry, but NACK.
>> >
>> > You have not explained why doing this correctly via ethtool -S cannot
>> > be done. debugfs is a vendor crap way of doing this.
>>
>> The debugfs interface is pre-existing and **in addition** to the
>> standard Linux interfaces which are also provided. It is true that
>> the debugfs interface in this case doesn't provide much additional
>> value apart from having the counter listed next to the hardware-
>> specific hashtable keys. As the debugfs interface for now aims to
>> be as complete as possible, naturally there is some redundance of
>> things which can also be accessed using other (standard) interfaces.
>
> debugfs is by definition unstable. It is not ABI. Anything using it is
> expected to break in the near future when it changes its layout. It is
> also totally option, you cannot expect it to be mounted.
>
> I hope you don't have any user space code using it.
>
> Maybe i should submit a patch which just for the fun of it rearranged
> the order in debugfs and change the file name?

I believe that OpenWrt is still the main user of the PPE offloading
code, since most vendors of devices with these SoC still use different
out-of-tree implementations.
OpenWrt does not ship or contain any user space code that relies on
these debugfs files.

Whenever I'm debugging PPE related issues, I rely heavily on the debugfs
API, so keeping it as complete as possible is important to me as well.

Aside from that, exposing per-flow statistics (which is what this patch
does) via ethtool -S API makes no sense to me at all. I'm not aware of
any other offload capable driver that does this.

- Felix

2022-11-02 10:28:13

by Daniel Golle

[permalink] [raw]
Subject: Re: [PATCH v3] net: ethernet: mediatek: ppe: add support for flow accounting

On Wed, Nov 02, 2022 at 02:26:17AM +0100, Andrew Lunn wrote:
> On Wed, Nov 02, 2022 at 01:03:08AM +0000, Daniel Golle wrote:
> > Hi Andrew,
> >
> > On Wed, Nov 02, 2022 at 01:51:21AM +0100, Andrew Lunn wrote:
> > > On Wed, Nov 02, 2022 at 12:42:40AM +0000, Daniel Golle wrote:
> > > > The PPE units found in MT7622 and newer support packet and byte
> > > > accounting of hw-offloaded flows. Add support for reading those
> > > > counters as found in MediaTek's SDK[1].
> > > >
> > > > [1]: https://git01.mediatek.com/plugins/gitiles/openwrt/feeds/mtk-openwrt-feeds/+/bc6a6a375c800dc2b80e1a325a2c732d1737df92
> > > > Signed-off-by: Daniel Golle <[email protected]>
> > >
> > > Sorry, but NACK.
> > >
> > > You have not explained why doing this correctly via ethtool -S cannot
> > > be done. debugfs is a vendor crap way of doing this.
> >
> > The debugfs interface is pre-existing and **in addition** to the
> > standard Linux interfaces which are also provided. It is true that
> > the debugfs interface in this case doesn't provide much additional
> > value apart from having the counter listed next to the hardware-
> > specific hashtable keys. As the debugfs interface for now aims to
> > be as complete as possible, naturally there is some redundance of
> > things which can also be accessed using other (standard) interfaces.
>
> debugfs is by definition unstable. It is not ABI. Anything using it is
> expected to break in the near future when it changes its layout. It is
> also totally option, you cannot expect it to be mounted.
>
> I hope you don't have any user space code using it.
>
> Maybe i should submit a patch which just for the fun of it rearranged
> the order in debugfs and change the file name?

Go ahead and have fun ;)
The debugfs interface is not used for anything except for the human
developer wanting to validate that PPE is working properly. It's
literally for debugging, and nothing else.

For productive use, you can access the counters by setting up an
nftables rule with a counter.