- 11 Jun, 2021 20 commits
-
-
Julian Wiedmann authored
Maintaining a pointer inside the aob's user-definable area is fragile and unnecessary. At this stage we only need it to overload the buffer's state field, and to access the buffer's TX queue. The first part is easily solved by tracking the aob's state within the aob itself. This also feels much cleaner and self-contained. For enabling the access to the associated TX queue, we can store the queue's index in the aob. Signed-off-by:
Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Julian Wiedmann authored
With commit 396c1004 ("s390/qdio: let driver manage the QAOB") a pending TX buffer now has access to its associated QAOB during TX completion processing. We can thus reduce the amount of work & state propagation that needs to be done by qeth_qdio_handle_aob(). Move all this logic into the respective TX completion paths. Doing so even allows us to determine more precise TX_NOTIFY_* values via qeth_compute_cq_notification(aob->aorc, ...). Signed-off-by:
-
Julian Wiedmann authored
Use a recently introduced helper to fill our ethtool stats strings. Signed-off-by:
-
Julian Wiedmann authored
We have one field to track _whether_ a cmd is active on a ccw device ('irq_pending'), and one to track _which_ cmd it is ('active_cmd'). Get rid of the irq_pending field, by testing active_cmd for NULL. Signed-off-by: -
Julian Wiedmann authored
Set scan_threshold = 0 to opt out from the qdio layer's internal tasklet & timer mechanism for TX completions, and replace it with the TX NAPI infrastructure that qeth already uses for IQD devices. This avoids the fragile logic in qdio_check_output_queue(), enables tighter integration and gives us more tuning options via ethtool in the future. For now we continue to apply the same policy as the qdio layer: scan for completions if 32 TX buffers are in use, or after 1 sec. A re-scan is done after 10 sec, but only if no TX interrupt is pending. With scan_threshold = 0 we no longer get TX completion scans from within qdio_get_next_buffers(). So trigger these manually in qeth_poll() and in the RX path switch to the equivalent qdio_inspect_queue(). Signed-off-by:
-
Julian Wiedmann authored
While the qdio layer already tracks the number of HW interrupts for a device, there's value in understanding how many of them have been raised due to our TX completion logic. Signed-off-by:
-
David S. Miller authored
Vladimir Oltean says: ==================== DSA tagging driver for NXP SJA1110 This series adds support for tagging data and control packets on the new NXP SJA1110 switch (supported by the sja1105 driver). Up to this point it used the sja1105 driver, which allowed it to send data packets, but not PDUs as those required by STP and PTP. To accommodate this new tagger which has both a header and a trailer, we need to refactor the entire DSA tagging scheme, to replace the "overhead" concept with separate "needed_headroom" and "needed_tailroom" concepts, so that SJA1110 can declare its need for both. There is also some consolidation work for the receive path of tag_8021q and its callers (sja1105 and ocelot-8021q). Changes in v3: Rebase in front of the "Port the SJA1105 DSA driver to XPCS" series which seems to have stalled for now. Changes in v2: Export the dsa_8021q_rcv and sja1110_process_meta_tstamp symbols to avoid build errors as modules. ==================== Signed-off-by: -
Vladimir Oltean authored
The TX timestamping procedure for SJA1105 is a bit unconventional because the transmit procedure itself is unconventional. Control packets (and therefore PTP as well) are transmitted to a specific port in SJA1105 using "management routes" which must be written over SPI to the switch. These are one-shot rules that match by destination MAC address on traffic coming from the CPU port, and select the precise destination port for that packet. So to transmit a packet from NET_TX softirq context, we actually need to defer to a process context so that we can perform that SPI write before we send the packet. The DSA master dev_queue_xmit() runs in process context, and we poll until the switch confirms it took the TX timestamp, then we annotate the skb clone with that TX timestamp. This is why the sja1105 driver does not need an skb queue for TX timestamping. But the SJA1110 is a bit (not much!) more conventional, and you can request 2-step TX timestamping through the DSA header, as well as give the switch a cookie (timestamp ID) which it will give back to you when it has the timestamp. So now we do need a queue for keeping the skb clones until their TX timestamps become available. The interesting part is that the metadata frames from SJA1105 haven't disappeared completely. On SJA1105 they were used as follow-ups which contained RX timestamps, but on SJA1110 they are actually TX completion packets, which contain a variable (up to 32) array of timestamps. Why an array? Because: - not only is the TX timestamp on the egress port being communicated, but also the RX timestamp on the CPU port. Nice, but we don't care about that, so we ignore it. - because a packet could be multicast to multiple egress ports, each port takes its own timestamp, and the TX completion packet contains the individual timestamps on each port. This is unconventional because switches typically have a timestamping FIFO and raise an interrupt, but this one doesn't. So the tagger needs to detect and parse meta frames, and call into the main switch driver, which pairs the timestamps with the skbs in the TX timestamping queue which are waiting for one. Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by:
-
Vladimir Oltean authored
This is really easy, since the full RX timestamp is in the DSA trailer and the tagger code transfers it to SJA1105_SKB_CB(skb)->tstamp, we just need to move it to the skb shared info region. This is as opposed to SJA1105, where the RX timestamp was received in a meta frame (so there needed to be a state machine to pair the 2 packets) and the timestamp was partial (so the packet, once matched with its timestamp, needed to be added to an RX timestamping queue where the PTP aux worker would reconstruct that timestamp). Signed-off-by:
-
Vladimir Oltean authored
The SJA1110 has improved a few things compared to SJA1105: - To send a control packet from the host port with SJA1105, one needed to program a one-shot "management route" over SPI. This is no longer true with SJA1110, you can actually send "in-band control extensions" in the packets sent by DSA, these are in fact DSA tags which contain the destination port and switch ID. - When receiving a control packet from the switch with SJA1105, the source port and switch ID were written in bytes 3 and 4 of the destination MAC address of the frame (which was a very poor shot at a DSA header). If the control packet also had an RX timestamp, that timestamp was sent in an actual follow-up packet, so there were reordering concerns on multi-core/multi-queue DSA masters, where the metadata frame with the RX timestamp might get processed before the actual packet to which that timestamp belonged (there is no way to pair a packet to its timestamp other than the order in which they were received). On SJA1110, this is no longer true, control packets have the source port, switch ID and timestamp all in the DSA tags. - Timestamps from the switch were partial: to get a 64-bit timestamp as required by PTP stacks, one would need to take the partial 24-bit or 32-bit timestamp from the packet, then read the current PTP time very quickly, and then patch in the high bits of the current PTP time into the captured partial timestamp, to reconstruct what the full 64-bit timestamp must have been. That is awful because packet processing is done in NAPI context, but reading the current PTP time is done over SPI and therefore needs sleepable context. But it also aggravated a few things: - Not only is there a DSA header in SJA1110, but there is a DSA trailer in fact, too. So DSA needs to be extended to support taggers which have both a header and a trailer. Very unconventional - my understanding is that the trailer exists because the timestamps couldn't be prepared in time for putting them in the header area. - Like SJA1105, not all packets sent to the CPU have the DSA tag added to them, only control packets do: * the ones which match the destination MAC filters/traps in MAC_FLTRES1 and MAC_FLTRES0 * the ones which match FDB entries which have TRAP or TAKETS bits set So we could in theory hack something up to request the switch to take timestamps for all packets that reach the CPU, and those would be DSA-tagged and contain the source port / switch ID by virtue of the fact that there needs to be a timestamp trailer provided. BUT: - The SJA1110 does not parse its own DSA tags in a way that is useful for routing in cross-chip topologies, a la Marvell. And the sja1105 driver already supports cross-chip bridging from the SJA1105 days. It does that by automatically setting up the DSA links as VLAN trunks which contain all the necessary tag_8021q RX VLANs that must be communicated between the switches that span the same bridge. So when using tag_8021q on sja1105, it is possible to have 2 switches with ports sw0p0, sw0p1, sw1p0, sw1p1, and 2 VLAN-unaware bridges br0 and br1, and br0 can take sw0p0 and sw1p0, and br1 can take sw0p1 and sw1p1, and forwarding will happen according to the expected rules of the Linux bridge. We like that, and we don't want that to go away, so as a matter of fact, the SJA1110 tagger still needs to support tag_8021q. So the sja1110 tagger is a hybrid between tag_8021q for data packets, and the native hardware support for control packets. On RX, packets have a 13-byte trailer if they contain an RX timestamp. That trailer is padded in such a way that its byte 8 (the start of the "residence time" field - not parsed by Linux because we don't care) is aligned on a 16 byte boundary. So the padding has a variable length between 0 and 15 bytes. The DSA header contains the offset of the beginning of the padding relative to the beginning of the frame (and the end of the padding is obviously the end of the packet minus 13 bytes, the length of the trailer). So we discard it. Packets which don't have a trailer contain the source port and switch ID information in the header (they are "trap-to-host" packets). Packets which have a trailer contain the source port and switch ID in the trailer. On TX, the destination port mask and switch ID is always in the trailer, so we always need to say in the header that a trailer is present. The header needs a custom EtherType and this was chosen as 0xdadc, after 0xdada which is for Marvell and 0xdadb which is for VLANs in VLAN-unaware mode on SJA1105 (and SJA1110 in fact too). Because we use tag_8021q in concert with the native tagging protocol, control packets will have 2 DSA tags. Signed-off-by: -
Vladimir Oltean authored
In SJA1105, RX timestamps for packets sent to the CPU are transmitted in separate follow-up packets (metadata frames). These contain partial timestamps (24 or 32 bits) which are kept in SJA1105_SKB_CB(skb)->meta_tstamp. Thankfully, SJA1110 improved that, and the RX timestamps are now transmitted in-band with the actual packet, in the timestamp trailer. The RX timestamps are now full-width 64 bits. Because we process the RX DSA tags in the rcv() method in the tagger, but we would like to preserve the DSA code structure in that we populate the skb timestamp in the port_rxtstamp() call which only happens later, the implication is that we must somehow pass the 64-bit timestamp from the rcv() method all the way to port_rxtstamp(). We can use the skb->cb for that. Rename the meta_tstamp from struct sja1105_skb_cb from "meta_tstamp" to "tstamp", and increase its size to 64 bits. Signed-off-by:
-
Vladimir Oltean authored
The added value of this function is that it can deal with both the case where the VLAN header is in the skb head, as well as in the offload field. This is something I was not able to do using other functions in the network stack. Since both ocelot-8021q and sja1105 need to do the same stuff, let's make it a common service provided by tag_8021q. This is done as refactoring for the new SJA1110 tagger, which partly uses tag_8021q as well (just like SJA1105), and will be the third caller. Signed-off-by:
-
Vladimir Oltean authored
All users of tag_8021q select it in Kconfig, so shim functions are not needed because it is not possible for it to be disabled and its callers enabled. Signed-off-by:
-
Vladimir Oltean authored
This makes no sense and is not needed, it is probably a debugging leftover. Signed-off-by:
-
Vladimir Oltean authored
Some really really weird switches just couldn't decide whether to use a normal or a tail tagger, so they just did both. This creates problems for DSA, because we only have the concept of an 'overhead' which can be applied to the headroom or to the tailroom of the skb (like for example during the central TX reallocation procedure), depending on the value of bool tail_tag, but not to both. We need to generalize DSA to cater for these odd switches by transforming the 'overhead / tail_tag' pair into 'needed_headroom / needed_tailroom'. The DSA master's MTU is increased to account for both. The flow dissector code is modified such that it only calls the DSA adjustment callback if the tagger has a non-zero header length. Taggers are trivially modified to declare either needed_headroom or needed_tailroom, based on the tail_tag value that they currently declare. Signed-off-by:
-
Vladimir Oltean authored
On SJA1105, there is support for a cascade port which is presumably connected to a downstream SJA1105 switch. The upstream one does not take PTP timestamps for packets received on this port, presumably because the downstream switch already did (and for PTP, it only makes sense for the leaf nodes in a DSA switch tree to do that). I haven't been able to validate that feature in a fully assembled setup, so I am disabling the feature by setting the cascade port to an unused port value (ds->num_ports). In SJA1110, multiple cascade ports are supported, and CASC_PORT became a bit mask from a port number. So when CASC_PORT is set to ds->num_ports (which is 11 on SJA1110), it is actually set to 0b1011, so ports 3, 1 and 0 are configured as cascade ports and we cannot take RX timestamps on them. So we need to introduce a check for SJA1110 and set things differently (to zero there), so that the cascading feature is properly disabled and RX timestamps can be taken on all ports. Signed-off-by:
-
Vladimir Oltean authored
As opposed to SJA1105 where there are parts with TTEthernet and parts without, in SJA1110 all parts support it, but it must be enabled in the static config. So enable it unconditionally. We use it for the tc-taprio and tc-gate offload. Signed-off-by:
-
David S. Miller authored
Guangbin Huang says: ==================== net: hns3: add support for PTP This series adds PTP support for the HNS3 ethernet driver. change log: V1 -> V2: 1. use spinlock to prevent concurrency 2. add the handling when ptp_clock_register() returns NULL ==================== Signed-off-by: -
Huazhong Tan authored
Add a debugfs interface for dumping ptp information, which is helpful for debugging. Signed-off-by:
Huazhong Tan <tanhuazhong@huawei.com> Signed-off-by:
Yufeng Mo <moyufeng@huawei.com> Signed-off-by:
Guangbin Huang <huangguangbin2@huawei.com> Signed-off-by:
-
Huazhong Tan authored
Adds PTP support for HNS3 ethernet driver. Signed-off-by:
-
- 10 Jun, 2021 20 commits
-
-
David S. Miller authored
Alex Elder says: ==================== net: ipa: memory region rework, part 2 This is the second portion of a set of patches updating the IPA memory region code. In this portion (part 2), the focus is on adjusting the code so that it no longer assumes the memory region descriptor array is indexed by the region identifier. This brings with it some related cleanup. Three loops are changed so their loop index variable is an unsigned rather than an enumerated type. A set of functions is changed so a region identifier (rather than a memory region descriptor pointer) is passed as argument, to simplify their call sites. This isn't entirely related or required, but I think it improves the code. A validation function for filter and route table memory regions is changed to take memory region IDs, rather than determining which region to validate based on a set of Boolean flags. Finally, ipa_mem_find() is created to abstract getting a memory descriptor based on its ID, and it is used everywhere rather than indexing the array. With that implemented, all of the memory regions can be defined by arrays of entries defined without providing index designators. ==================== Signed-off-by: -
Alex Elder authored
Finally the code handles the IPA memory region array in the configuration data without assuming it is indexed by region ID. Get rid of the array index designators where these arrays are initialized. As a result, there's no more need to define an explicitly undefined memory region ID, so get rid of that. Change ipa_mem_find() so it no longer assumes the ipa->mem[] array is indexed by memory region ID. Instead, have it search the array for the entry having the requested memory ID, and return the address of the descriptor if found. Otherwise return NULL. Stop allowing memory regions to be defined with zero size and zero canary value. Check for this condition in ipa_mem_valid_one(). As a result, it is not necessary to check for this case in ipa_mem_config(). Finally, there is no need for IPA_MEM_UNDEFINED to be defined any more, so get rid of it. Signed-off-by:
Alex Elder <elder@linaro.org> Signed-off-by:
-
Alex Elder authored
Introduce a new function that abstracts finding information about a region in IPA-local memory, given its memory region ID. For now it simply uses the region ID as an index into the IPA memory array. If the region is not defined, ipa_mem_find() returns a null pointer. Update all code that accesses the ipa->mem[] array directly to use ipa_mem_find() instead. The return value must be checked for null when optional memory regions are sought. Signed-off-by:
-
Alex Elder authored
Stop passing most of the Boolean flags to ipa_table_valid_one(), and just pass a memory region ID to it instead. We still need to indicate whether we're operating on a routing or filter table. Signed-off-by:
-
Alex Elder authored
Pass a memory region ID rather than the address of a memory region descriptor to ipa_table_reset_add() to simplify callers. Similarly, pass memory region IDs to ipa_table_init_add(). Signed-off-by:
-
Alex Elder authored
Pass a memory region ID rather than the address of a memory region descriptor to ipa_mem_zero_region_add() to simplify callers. Signed-off-by:
-
Alex Elder authored
Pass a memory region ID rather than the address of a memory region descriptor to ipa_filter_reset_table(), to simplify callers. We can eliminate the check for a zero region size in this function because ipa_table_reset_add() checks that before adding anything to the transaction. Note that here and in subsequent commits there is no need to check whether a memory region exists, because we will have already verified that during initialization. Signed-off-by:
-
Alex Elder authored
Do some general cleanup in ipa_cmd_header_valid(): - Delay assigning the mem variable until just before it's used. - Assign the maximum offset and size values together. - Improve comments explaining the single range of memory being made up of a modem portion and an AP portion. - Record the offset of the combined range in a local variable. - Do the initial size assignment right after assigning the offset. Signed-off-by: -
Alex Elder authored
Change ipa_mem_valid() to iterate over the entries using a u32 index variable rather than using a memory region ID. Use the ID found inside the memory descriptor rather than the loop index. Change ipa_mem_size_valid() to iterate over the entries but without assuming the array index is the memory region ID. "Empty" entries will have zero size; and we'll temporarily assume such entries have zero offset as well (they all do, currently). Similarly, don't assume the mem[] array is indexed by ID in ipa_mem_config(). There, "empty" entries will have a zero canary count, so no special assumptions are needed to handle them correctly. Signed-off-by:
-
Cristobal Forno authored
Allow the device to be initialized at a later time if it is not available at boot. The device will be allowed to probe but will be given a "down" state. After completing device probe and registering the net device, the driver will await an interrupt signal from its partner device, indicating that it is ready for boot. The driver will schedule a work event to perform the necessary procedure and begin operation. Co-developed-by:
Thomas Falcon <tlfalcon@linux.ibm.com> Signed-off-by:
Cristobal Forno <cforno12@linux.ibm.com> Acked-by:
Lijun Pan <lijunp213@gmail.com> Reviewed-by:
Dany Madden <drt@linux.ibm.com> Signed-off-by:
-
David S. Miller authored
Vadym Kochan says: ==================== net: marvell: prestera: add LAG support The following features are supported: - LAG basic operations - create/delete LAG - add/remove a member to LAG - enable/disable member in LAG - LAG Bridge support - LAG VLAN support - LAG FDB support Limitations: - Only HASH lag tx type is supported - The Hash parameters are not configurable. They are applied during the LAG creation stage. - Enslaving a port to the LAG device that already has an upper device is not supported. Changes extracted from: https://lkml.org/lkml/2021/2/3/877 and marked with "v2". v2: There are 2 additional preparation patches which simplifies the netdev topology handling. 1) Initialize 'lag' with NULL in prestera_lag_create() [suggested by Vladimir Oltean] 2) Use -ENOSPC in prestera_lag_port_add() if max lag [suggested by Vladimir Oltean] numbers were reached. 3) Do not propagate netdev events to prestera_switchdev [suggested by Vladimir Oltean] but call bridge specific funcs. It simplifies the code. 4) Check on info->link_up in prestera_netdev_port_lower_event() [suggested by Vladimir Oltean] 5) Return -EOPNOTSUPP in prestera_netdev_port_event() in case [suggested by Vladimir Oltean] LAG hashing mode is not supported. 6) Do not pass "lower" netdev to bridge join/leave functions. [suggested by Vladimir Oltean] It is not need as offloading settings applied on particular physical port. It requires to do extra upper dev lookup in case port is in the LAG which is in the bridge on vlans add/del. ==================== Signed-off-by: -
Serhiy Boiko authored
The following features are supported: - LAG basic operations - create/delete LAG - add/remove a member to LAG - enable/disable member in LAG - LAG Bridge support - LAG VLAN support - LAG FDB support Limitations: - Only HASH lag tx type is supported - The Hash parameters are not configurable. They are applied during the LAG creation stage. - Enslaving a port to the LAG device that already has an upper device is not supported. Co-developed-by:Andrii Savka <andrii.savka@plvision.eu> Signed-off-by:
Serhiy Boiko <serhiy.boiko@plvision.eu> Co-developed-by:
Vadym Kochan <vkochan@marvell.com> Signed-off-by:
-
Vadym Kochan authored
Replace prestera_bridge_port_event(...) by prestera_bridge_port_join(...) and prestera_bridge_port_leave(). It simplifies the code by reading netdev event specific handling only once in prestera_main.c Signed-off-by:
-
Vadym Kochan authored
Move handling of PRECHANGEUPPER event from prestera_switchdev to prestera_main which is responsible for basic netdev events handling and routing them to related module. Signed-off-by:
-
Tan Zhongjun authored
The platform_get_irq() prints error message telling that interrupt is missing,hence there is no need to duplicated that message in the drivers. Signed-off-by:
Tan Zhongjun <tanzhongjun@yulong.com> Signed-off-by:
-
Baokun Li authored
Add description for `tfrc_invert_loss_event_rate` to fix the W=1 warnings: net/dccp/ccids/lib/tfrc_equation.c:695: warning: Function parameter or member 'loss_event_rate' not described in 'tfrc_invert_loss_event_rate' Signed-off-by:
Baokun Li <libaokun1@huawei.com> Reviewed-by:
Richard Sailer <richard_siegfried@systemli.org> Signed-off-by:
-
Wang Hai authored
Convert list_for_each() to list_for_each_entry() where applicable. This simplifies the code. Reported-by:
Hulk Robot <hulkci@huawei.com> Signed-off-by:
Wang Hai <wanghai38@huawei.com> Signed-off-by:
-
Wang Hai authored
Convert list_for_each() to list_for_each_entry() where applicable. This simplifies the code. Reported-by:
-
Wang Hai authored
Convert list_for_each() to list_for_each_entry() where applicable. This simplifies the code. Reported-by:
-
Yang Yingliang authored
Use devm_platform_get_and_ioremap_resource() to simplify code. Signed-off-by:
Yang Yingliang <yangyingliang@huawei.com> Signed-off-by:
-
