cxgb4: Add support for dynamic allocation of resources for ULD
Add a new commmon infrastructure to allocate reosurces dynamically to
Upper layer driver's(ULD) when they register with cxgb4 driver and free
them during unregistering. All the queues and the interrupts for
them will be allocated during ULD probe only and freed during remove.
Signed-off-by: Atul Gupta <atul.gupta@chelsio.com>
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 2bb804c..85e30f1 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -223,8 +223,8 @@
static struct dentry *cxgb4_debugfs_root;
-static LIST_HEAD(adapter_list);
-static DEFINE_MUTEX(uld_mutex);
+LIST_HEAD(adapter_list);
+DEFINE_MUTEX(uld_mutex);
/* Adapter list to be accessed from atomic context */
static LIST_HEAD(adap_rcu_list);
static DEFINE_SPINLOCK(adap_rcu_lock);
@@ -1066,20 +1066,20 @@
*/
static int setup_sge_queues(struct adapter *adap)
{
- int err, msi_idx, i, j;
+ int err, i, j;
struct sge *s = &adap->sge;
bitmap_zero(s->starving_fl, s->egr_sz);
bitmap_zero(s->txq_maperr, s->egr_sz);
if (adap->flags & USING_MSIX)
- msi_idx = 1; /* vector 0 is for non-queue interrupts */
+ adap->msi_idx = 1; /* vector 0 is for non-queue interrupts */
else {
err = t4_sge_alloc_rxq(adap, &s->intrq, false, adap->port[0], 0,
NULL, NULL, NULL, -1);
if (err)
return err;
- msi_idx = -((int)s->intrq.abs_id + 1);
+ adap->msi_idx = -((int)s->intrq.abs_id + 1);
}
/* NOTE: If you add/delete any Ingress/Egress Queue allocations in here,
@@ -1096,7 +1096,7 @@
* new/deleted queues.
*/
err = t4_sge_alloc_rxq(adap, &s->fw_evtq, true, adap->port[0],
- msi_idx, NULL, fwevtq_handler, NULL, -1);
+ adap->msi_idx, NULL, fwevtq_handler, NULL, -1);
if (err) {
freeout: t4_free_sge_resources(adap);
return err;
@@ -1109,10 +1109,10 @@
struct sge_eth_txq *t = &s->ethtxq[pi->first_qset];
for (j = 0; j < pi->nqsets; j++, q++) {
- if (msi_idx > 0)
- msi_idx++;
+ if (adap->msi_idx > 0)
+ adap->msi_idx++;
err = t4_sge_alloc_rxq(adap, &q->rspq, false, dev,
- msi_idx, &q->fl,
+ adap->msi_idx, &q->fl,
t4_ethrx_handler,
NULL,
t4_get_mps_bg_map(adap,
@@ -1141,11 +1141,11 @@
}
#define ALLOC_OFLD_RXQS(firstq, nq, per_chan, ids, lro) do { \
- err = alloc_ofld_rxqs(adap, firstq, nq, per_chan, msi_idx, ids, lro); \
+ err = alloc_ofld_rxqs(adap, firstq, nq, per_chan, adap->msi_idx, ids, lro); \
if (err) \
goto freeout; \
- if (msi_idx > 0) \
- msi_idx += nq; \
+ if (adap->msi_idx > 0) \
+ adap->msi_idx += nq; \
} while (0)
ALLOC_OFLD_RXQS(s->iscsirxq, s->iscsiqsets, j, s->iscsi_rxq, false);
@@ -2565,6 +2565,12 @@
CXGB4_STATE_DETACH);
adap->uld_handle[i] = NULL;
}
+ for (i = 0; i < CXGB4_PCI_ULD_MAX; i++)
+ if (adap->uld && adap->uld[i].handle) {
+ adap->uld[i].state_change(adap->uld[i].handle,
+ CXGB4_STATE_DETACH);
+ adap->uld[i].handle = NULL;
+ }
if (netevent_registered && list_empty(&adapter_list)) {
unregister_netevent_notifier(&cxgb4_netevent_nb);
netevent_registered = false;
@@ -2584,6 +2590,10 @@
for (i = 0; i < CXGB4_ULD_MAX; i++)
if (adap->uld_handle[i])
ulds[i].state_change(adap->uld_handle[i], new_state);
+ for (i = 0; i < CXGB4_PCI_ULD_MAX; i++)
+ if (adap->uld && adap->uld[i].handle)
+ adap->uld[i].state_change(adap->uld[i].handle,
+ new_state);
mutex_unlock(&uld_mutex);
}
@@ -4170,6 +4180,11 @@
adap->vres.iscsi.start = val[0];
adap->vres.iscsi.size = val[1] - val[0] + 1;
}
+ if (caps_cmd.cryptocaps) {
+ /* Should query params here...TODO */
+ adap->params.crypto |= ULP_CRYPTO_LOOKASIDE;
+ adap->num_uld += 1;
+ }
#undef FW_PARAM_PFVF
#undef FW_PARAM_DEV
@@ -4351,16 +4366,6 @@
(lc->supported & FW_PORT_CAP_SPEED_40G) != 0;
}
-static inline void init_rspq(struct adapter *adap, struct sge_rspq *q,
- unsigned int us, unsigned int cnt,
- unsigned int size, unsigned int iqe_size)
-{
- q->adap = adap;
- cxgb4_set_rspq_intr_params(q, us, cnt);
- q->iqe_len = iqe_size;
- q->size = size;
-}
-
/*
* Perform default configuration of DMA queues depending on the number and type
* of ports we found and the number of available CPUs. Most settings can be
@@ -4375,6 +4380,15 @@
#endif
int ciq_size;
+ /* Reduce memory usage in kdump environment, disable all offload.
+ */
+ if (is_kdump_kernel()) {
+ adap->params.offload = 0;
+ adap->params.crypto = 0;
+ } else if (adap->num_uld && uld_mem_alloc(adap)) {
+ adap->params.crypto = 0;
+ }
+
for_each_port(adap, i)
n10g += is_x_10g_port(&adap2pinfo(adap, i)->link_cfg);
#ifdef CONFIG_CHELSIO_T4_DCB
@@ -4405,11 +4419,6 @@
if (q10g > netif_get_num_default_rss_queues())
q10g = netif_get_num_default_rss_queues();
- /* Reduce memory usage in kdump environment, disable all offload.
- */
- if (is_kdump_kernel())
- adap->params.offload = 0;
-
for_each_port(adap, i) {
struct port_info *pi = adap2pinfo(adap, i);
@@ -4538,23 +4547,58 @@
}
}
+static int get_msix_info(struct adapter *adap)
+{
+ struct uld_msix_info *msix_info;
+ int max_ingq = (MAX_OFLD_QSETS * adap->num_uld);
+
+ msix_info = kcalloc(max_ingq, sizeof(*msix_info), GFP_KERNEL);
+ if (!msix_info)
+ return -ENOMEM;
+
+ adap->msix_bmap_ulds.msix_bmap = kcalloc(BITS_TO_LONGS(max_ingq),
+ sizeof(long), GFP_KERNEL);
+ if (!adap->msix_bmap_ulds.msix_bmap) {
+ kfree(msix_info);
+ return -ENOMEM;
+ }
+ spin_lock_init(&adap->msix_bmap_ulds.lock);
+ adap->msix_info_ulds = msix_info;
+ return 0;
+}
+
+static void free_msix_info(struct adapter *adap)
+{
+ if (!adap->num_uld)
+ return;
+
+ kfree(adap->msix_info_ulds);
+ kfree(adap->msix_bmap_ulds.msix_bmap);
+}
+
/* 2 MSI-X vectors needed for the FW queue and non-data interrupts */
#define EXTRA_VECS 2
static int enable_msix(struct adapter *adap)
{
- int ofld_need = 0;
- int i, want, need, allocated;
+ int ofld_need = 0, uld_need = 0;
+ int i, j, want, need, allocated;
struct sge *s = &adap->sge;
unsigned int nchan = adap->params.nports;
struct msix_entry *entries;
+ int max_ingq = MAX_INGQ;
- entries = kmalloc(sizeof(*entries) * (MAX_INGQ + 1),
+ max_ingq += (MAX_OFLD_QSETS * adap->num_uld);
+ entries = kmalloc(sizeof(*entries) * (max_ingq + 1),
GFP_KERNEL);
if (!entries)
return -ENOMEM;
- for (i = 0; i < MAX_INGQ + 1; ++i)
+ /* map for msix */
+ if (is_pci_uld(adap) && get_msix_info(adap))
+ adap->params.crypto = 0;
+
+ for (i = 0; i < max_ingq + 1; ++i)
entries[i].entry = i;
want = s->max_ethqsets + EXTRA_VECS;
@@ -4567,13 +4611,17 @@
else
ofld_need = 4 * nchan;
}
+ if (is_pci_uld(adap)) {
+ want += netif_get_num_default_rss_queues() * nchan;
+ uld_need = nchan;
+ }
#ifdef CONFIG_CHELSIO_T4_DCB
/* For Data Center Bridging we need 8 Ethernet TX Priority Queues for
* each port.
*/
- need = 8 * adap->params.nports + EXTRA_VECS + ofld_need;
+ need = 8 * adap->params.nports + EXTRA_VECS + ofld_need + uld_need;
#else
- need = adap->params.nports + EXTRA_VECS + ofld_need;
+ need = adap->params.nports + EXTRA_VECS + ofld_need + uld_need;
#endif
allocated = pci_enable_msix_range(adap->pdev, entries, need, want);
if (allocated < 0) {
@@ -4587,12 +4635,20 @@
* Every group gets its minimum requirement and NIC gets top
* priority for leftovers.
*/
- i = allocated - EXTRA_VECS - ofld_need;
+ i = allocated - EXTRA_VECS - ofld_need - uld_need;
if (i < s->max_ethqsets) {
s->max_ethqsets = i;
if (i < s->ethqsets)
reduce_ethqs(adap, i);
}
+ if (is_pci_uld(adap)) {
+ if (allocated < want)
+ s->nqs_per_uld = nchan;
+ else
+ s->nqs_per_uld = netif_get_num_default_rss_queues() *
+ nchan;
+ }
+
if (is_offload(adap)) {
if (allocated < want) {
s->rdmaqs = nchan;
@@ -4604,16 +4660,24 @@
/* leftovers go to OFLD */
i = allocated - EXTRA_VECS - s->max_ethqsets -
- s->rdmaqs - s->rdmaciqs - s->niscsitq;
+ s->rdmaqs - s->rdmaciqs - s->niscsitq;
+ if (is_pci_uld(adap))
+ i -= s->nqs_per_uld * adap->num_uld;
s->iscsiqsets = (i / nchan) * nchan; /* round down */
}
- for (i = 0; i < allocated; ++i)
+
+ for (i = 0; i < (allocated - (s->nqs_per_uld * adap->num_uld)); ++i)
adap->msix_info[i].vec = entries[i].vector;
+ if (is_pci_uld(adap)) {
+ for (j = 0 ; i < allocated; ++i, j++)
+ adap->msix_info_ulds[j].vec = entries[i].vector;
+ adap->msix_bmap_ulds.mapsize = j;
+ }
dev_info(adap->pdev_dev, "%d MSI-X vectors allocated, "
- "nic %d iscsi %d rdma cpl %d rdma ciq %d\n",
+ "nic %d iscsi %d rdma cpl %d rdma ciq %d uld %d\n",
allocated, s->max_ethqsets, s->iscsiqsets, s->rdmaqs,
- s->rdmaciqs);
+ s->rdmaciqs, s->nqs_per_uld);
kfree(entries);
return 0;
@@ -5215,8 +5279,11 @@
/* See what interrupts we'll be using */
if (msi > 1 && enable_msix(adapter) == 0)
adapter->flags |= USING_MSIX;
- else if (msi > 0 && pci_enable_msi(pdev) == 0)
+ else if (msi > 0 && pci_enable_msi(pdev) == 0) {
adapter->flags |= USING_MSI;
+ if (msi > 1)
+ free_msix_info(adapter);
+ }
/* check for PCI Express bandwidth capabiltites */
cxgb4_check_pcie_caps(adapter);
@@ -5332,6 +5399,10 @@
out_free_dev:
free_some_resources(adapter);
+ if (adapter->flags & USING_MSIX)
+ free_msix_info(adapter);
+ if (adapter->num_uld)
+ uld_mem_free(adapter);
out_unmap_bar:
if (!is_t4(adapter->params.chip))
iounmap(adapter->bar2);
@@ -5393,6 +5464,10 @@
if (adapter->flags & FULL_INIT_DONE)
cxgb_down(adapter);
+ if (adapter->flags & USING_MSIX)
+ free_msix_info(adapter);
+ if (adapter->num_uld)
+ uld_mem_free(adapter);
free_some_resources(adapter);
#if IS_ENABLED(CONFIG_IPV6)
t4_cleanup_clip_tbl(adapter);