pds_core: Various fixes
Subsystems (3)
Brett Creeley
maintainercover pds_core: Various fixes
This series includes the following changes:
There can be many users of the pds_core's adminq. This includes pds_core's uses and any clients that depend on it. When the pds_core device goes through a reset for any reason the adminq is freed and reconfigured. There are some gaps in the current implementation that will cause crashes during reset if any of the previously mentioned users of the adminq attempt to use it after it's been freed.
Issues around how resets are handled, specifically regarding the driver's error handlers.
Originally these patches were aimed at net-next, but it was requested to push the fixes patches to net. The original patches can be found here:
https://lore.kernel.org/netdev/[email protected]/
Also, the Reviewed-by tags were left in place from net-next reviews as the patches didn't change.
Brett Creeley (6):
pds_core: Prevent health thread from running during reset/remove
pds_core: Cancel AQ work on teardown
pds_core: Use struct pdsc for the pdsc_adminq_isr private data
pds_core: Prevent race issues involving the adminq
pds_core: Clear BARs on reset
pds_core: Rework teardown/setup flow to be more common
drivers/net/ethernet/amd/pds_core/adminq.c | 64 +++++++++++++++------ drivers/net/ethernet/amd/pds_core/core.c | 46 +++++++++++---- drivers/net/ethernet/amd/pds_core/core.h | 2 +- drivers/net/ethernet/amd/pds_core/debugfs.c | 4 ++ drivers/net/ethernet/amd/pds_core/dev.c | 16 +++--- drivers/net/ethernet/amd/pds_core/devlink.c | 3 +- drivers/net/ethernet/amd/pds_core/fw.c | 3 + drivers/net/ethernet/amd/pds_core/main.c | 26 +++++++-- 8 files changed, 124 insertions(+), 40 deletions(-)
Signature · 1 line
2.17.1
Brett Creeley
maintainerPATCH 1/6 pds_core: Prevent health thread from running during reset/remove
The PCIe reset handlers can run at the same time as the health thread. This can cause the health thread to stomp on the PCIe reset. Fix this by preventing the health thread from running while a PCIe reset is happening.
As part of this use timer_shutdown_sync() during reset and remove to make sure the timer doesn't ever get rearmed.
--- drivers/net/ethernet/amd/pds_core/main.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-)
drivers/net/ethernet/amd/pds_core/main.c +17 -2
@@ -293,7 +293,7 @@ static int pdsc_init_pf(struct pdsc *pdsc) err_out_teardown:pdsc_teardown (pdsc,PDSC_TEARDOWN_REMOVING ); err_out_unmap_bars: -del_timer_sync (&pdsc->wdtimer); +timer_shutdown_sync (&pdsc->wdtimer);if (pdsc->wq)destroy_workqueue (pdsc->wq);mutex_destroy (&pdsc->config_lock);
@@ -420,7 +420,7 @@ static void pdsc_remove(struct pci_dev *pdev)*/ pdsc_sriov_configure (pdev,0 ); -del_timer_sync (&pdsc->wdtimer); +timer_shutdown_sync (&pdsc->wdtimer);if (pdsc->wq)destroy_workqueue (pdsc->wq);
@@ -445,10 +445,24 @@ static void pdsc_remove(struct pci_dev *pdev)devlink_free (dl); } +static void pdsc_stop_health_thread (struct pdsc *pdsc) +{ +timer_shutdown_sync (&pdsc->wdtimer); +if (pdsc->health_work.func) +cancel_work_sync (&pdsc->health_work); +} + +static void pdsc_restart_health_thread (struct pdsc *pdsc) +{ +timer_setup (&pdsc->wdtimer, pdsc_wdtimer_cb,0 ); +mod_timer (&pdsc->wdtimer, jiffies +1 ); +} +void pdsc_reset_prepare (struct pci_dev *pdev) {struct pdsc *pdsc =pci_get_drvdata (pdev); +pdsc_stop_health_thread (pdsc);pdsc_fw_down (pdsc);pci_free_irq_vectors (pdev);
@@ -486,6 +500,7 @@ void pdsc_reset_done(struct pci_dev *pdev) }pdsc_fw_up (pdsc); +pdsc_restart_health_thread (pdsc); }static const struct pci_error_handlers pdsc_err_handler = {
Signature · 1 line
2.17.1
Brett Creeley
maintainerPATCH 2/6 pds_core: Cancel AQ work on teardown
In reply to Brett CreeleyThere is a small window where pdsc_work_thread() calls pdsc_process_adminq() and pdsc_process_adminq() passes the PDSC_S_STOPPING_DRIVER check and starts to process adminq/notifyq work and then the driver starts a fw_down cycle. This could cause some undefined behavior if the notifyqcq/adminqcq are free'd while pdsc_process_adminq() is running. Use cancel_work_sync() on the adminqcq's work struct to make sure any pending work items are cancelled and any in progress work items are completed.
Also, make sure to not call cancel_work_sync() if the work item has not be initialized. Without this, traces will happen in cases where a reset fails and teardown is called again or if reset fails and the driver is removed.
--- drivers/net/ethernet/amd/pds_core/core.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/net/ethernet/amd/pds_core/core.c +2 -0
@@ -464,6 +464,8 @@ void pdsc_teardown(struct pdsc *pdsc, bool removing)if (!pdsc->pdev->is_virtfn)pdsc_devcmd_reset (pdsc); +if (pdsc->adminqcq.work.func) +cancel_work_sync (&pdsc->adminqcq.work);pdsc_qcq_free (pdsc, &pdsc->notifyqcq);pdsc_qcq_free (pdsc, &pdsc->adminqcq);
Signature · 1 line
2.17.1
Brett Creeley
maintainerPATCH 3/6 pds_core: Use struct pdsc for the pdsc_adminq_isr private data
In reply to Brett CreeleyThe initial design for the adminq interrupt was done based on client drivers having their own adminq and adminq interrupt. So, each client driver's adminq isr would use their specific adminqcq for the private data struct. For the time being the design has changed to only use a single adminq for all clients. So, instead use the struct pdsc for the private data to simplify things a bit.
This also has the benefit of not dereferencing the adminqcq to access the pdsc struct when the PDSC_S_STOPPING_DRIVER bit is set and the adminqcq has actually been cleared/freed.
--- drivers/net/ethernet/amd/pds_core/adminq.c | 5 +++-- drivers/net/ethernet/amd/pds_core/core.c | 2 +- 2 files changed, 4 insertions(+), 3 deletions(-)
drivers/net/ethernet/amd/pds_core/adminq.c +3 -2
@@ -135,8 +135,8 @@ void pdsc_work_thread(struct work_struct *work)irqreturn_t pdsc_adminq_isr (int irq,void *data) { -struct pdsc_qcq *qcq = data; -struct pdsc *pdsc = qcq->pdsc; +struct pdsc *pdsc = data; +struct pdsc_qcq *qcq;/* Don't process AdminQ when shutting down */ if (pdsc->state &BIT_ULL (PDSC_S_STOPPING_DRIVER )) {
@@ -145,6 +145,7 @@ irqreturn_t pdsc_adminq_isr(int irq, void *data)return IRQ_HANDLED ; } + qcq = &pdsc->adminqcq;queue_work (pdsc->wq, &qcq->work);pds_core_intr_mask (&pdsc->intr_ctrl[qcq->intx],PDS_CORE_INTR_MASK_CLEAR );
drivers/net/ethernet/amd/pds_core/core.c +1 -1
@@ -125,7 +125,7 @@ static int pdsc_qcq_intr_alloc(struct pdsc *pdsc, struct pdsc_qcq *qcq)snprintf (name,sizeof (name),"%s-%d-%s" ,PDS_CORE_DRV_NAME , pdsc->pdev->bus->number, qcq->q.name); - index =pdsc_intr_alloc (pdsc, name, pdsc_adminq_isr, qcq); + index =pdsc_intr_alloc (pdsc, name, pdsc_adminq_isr, pdsc);if (index <0 )return index; qcq->intx = index;
Signature · 1 line
2.17.1
Brett Creeley
maintainerPATCH 4/6 pds_core: Prevent race issues involving the adminq
In reply to Brett CreeleyThere are multiple paths that can result in using the pdsc's adminq.
[1] pdsc_adminq_isr and the resulting work from queue_work(),
i.e. pdsc_work_thread()->pdsc_process_adminq()
[2] pdsc_adminq_post()
When the device goes through reset via PCIe reset and/or a fw_down/fw_up cycle due to bad PCIe state or bad device state the adminq is destroyed and recreated.
A NULL pointer dereference can happen if [1] or [2] happens after the adminq is already destroyed.
In order to fix this, add some further state checks and implement reference counting for adminq uses. Reference counting was used because multiple threads can attempt to access the adminq at the same time via [1] or [2]. Additionally, multiple clients (i.e. pds-vfio-pci) can be using [2] at the same time.
The adminq_refcnt is initialized to 1 when the adminq has been allocated and is ready to use. Users/clients of the adminq (i.e. [1] and [2]) will increment the refcnt when they are using the adminq. When the driver goes into a fw_down cycle it will set the PDSC_S_FW_DEAD bit and then wait for the adminq_refcnt to hit 1. Setting the PDSC_S_FW_DEAD before waiting will prevent any further adminq_refcnt increments. Waiting for the adminq_refcnt to hit 1 allows for any current users of the adminq to finish before the driver frees the adminq. Once the adminq_refcnt hits 1 the driver clears the refcnt to signify that the adminq is deleted and cannot be used. On the fw_up cycle the driver will once again initialize the adminq_refcnt to 1 allowing the adminq to be used again.
--- drivers/net/ethernet/amd/pds_core/adminq.c | 31 +++++++++++++++++----- drivers/net/ethernet/amd/pds_core/core.c | 21 +++++++++++++++ drivers/net/ethernet/amd/pds_core/core.h | 1 + 3 files changed, 47 insertions(+), 6 deletions(-)
drivers/net/ethernet/amd/pds_core/adminq.c +25 -6
@@ -63,6 +63,15 @@ static int pdsc_process_notifyq(struct pdsc_qcq *qcq)return nq_work; } +static bool pdsc_adminq_inc_if_up (struct pdsc *pdsc) +{ +if (pdsc->state &BIT_ULL (PDSC_S_STOPPING_DRIVER ) || + pdsc->state &BIT_ULL (PDSC_S_FW_DEAD )) +return false; + +return refcount_inc_not_zero (&pdsc->adminq_refcnt); +} +void pdsc_process_adminq (struct pdsc_qcq *qcq) {union pds_core_adminq_comp *comp;
@@ -75,9 +84,9 @@ void pdsc_process_adminq(struct pdsc_qcq *qcq)int aq_work =0 ;int credits; -/* Don't process AdminQ when shutting down */ -if (pdsc->state &BIT_ULL (PDSC_S_STOPPING_DRIVER )) { -dev_err (pdsc->dev,"%s: called while PDSC_S_STOPPING_DRIVER\n" , +/* Don't process AdminQ when it's not up */ +if (!pdsc_adminq_inc_if_up (pdsc)) { +dev_err (pdsc->dev,"%s: called while adminq is unavailable\n" , __func__);return ; }
@@ -124,6 +133,7 @@ void pdsc_process_adminq(struct pdsc_qcq *qcq)pds_core_intr_credits (&pdsc->intr_ctrl[qcq->intx], credits,PDS_CORE_INTR_CRED_REARM ); +refcount_dec (&pdsc->adminq_refcnt); }void pdsc_work_thread (struct work_struct *work)
@@ -138,9 +148,9 @@ irqreturn_t pdsc_adminq_isr(int irq, void *data)struct pdsc *pdsc = data;struct pdsc_qcq *qcq; -/* Don't process AdminQ when shutting down */ -if (pdsc->state &BIT_ULL (PDSC_S_STOPPING_DRIVER )) { -dev_err (pdsc->dev,"%s: called while PDSC_S_STOPPING_DRIVER\n" , +/* Don't process AdminQ when it's not up */ +if (!pdsc_adminq_inc_if_up (pdsc)) { +dev_err (pdsc->dev,"%s: called while adminq is unavailable\n" , __func__);return IRQ_HANDLED ; }
@@ -148,6 +158,7 @@ irqreturn_t pdsc_adminq_isr(int irq, void *data) qcq = &pdsc->adminqcq;queue_work (pdsc->wq, &qcq->work);pds_core_intr_mask (&pdsc->intr_ctrl[qcq->intx],PDS_CORE_INTR_MASK_CLEAR ); +refcount_dec (&pdsc->adminq_refcnt);return IRQ_HANDLED ; }
@@ -231,6 +242,12 @@ int pdsc_adminq_post(struct pdsc *pdsc,int err =0 ;int index; +if (!pdsc_adminq_inc_if_up (pdsc)) { +dev_dbg (pdsc->dev,"%s: preventing adminq cmd %u\n" , + __func__, cmd->opcode); +return -ENXIO ; + } + wc.qcq = &pdsc->adminqcq; index =__pdsc_adminq_post (pdsc, &pdsc->adminqcq, cmd, comp, &wc);if (index <0 ) {
@@ -286,6 +303,8 @@ int pdsc_adminq_post(struct pdsc *pdsc,queue_work (pdsc->wq, &pdsc->health_work); } +refcount_dec (&pdsc->adminq_refcnt); +return err; }EXPORT_SYMBOL_GPL (pdsc_adminq_post);
drivers/net/ethernet/amd/pds_core/core.c +21 -0
@@ -450,6 +450,7 @@ int pdsc_setup(struct pdsc *pdsc, bool init)pdsc_debugfs_add_viftype (pdsc); } +refcount_set (&pdsc->adminq_refcnt,1 );clear_bit (PDSC_S_FW_DEAD , &pdsc->state);return 0 ;
@@ -514,6 +515,24 @@ void pdsc_stop(struct pdsc *pdsc) PDS_CORE_INTR_MASK_SET); } +static void pdsc_adminq_wait_and_dec_once_unused (struct pdsc *pdsc) +{ +/* The driver initializes the adminq_refcnt to 1 when the adminq is +* allocated and ready for use. Other users/requesters will increment +* the refcnt while in use. If the refcnt is down to 1 then the adminq +* is not in use and the refcnt can be cleared and adminq freed. Before +* calling this function the driver will set PDSC_S_FW_DEAD, which +* prevent subsequent attempts to use the adminq and increment the +* refcnt to fail. This guarantees that this function will eventually +* exit. +*/ +while (!refcount_dec_if_one (&pdsc->adminq_refcnt)) { +dev_dbg_ratelimited (pdsc->dev,"%s: adminq in use\n" , + __func__); +cpu_relax (); + } +} +void pdsc_fw_down (struct pdsc *pdsc) {union pds_core_notifyq_comp reset_event = {
@@ -529,6 +548,8 @@ void pdsc_fw_down(struct pdsc *pdsc)if (pdsc->pdev->is_virtfn)return ; +pdsc_adminq_wait_and_dec_once_unused (pdsc); +/* Notify clients of fw_down */ if (pdsc->fw_reporter)devlink_health_report (pdsc->fw_reporter,"FW down reported" , pdsc);
drivers/net/ethernet/amd/pds_core/core.h +1 -0
@@ -184,6 +184,7 @@ struct pdsc {struct mutex devcmd_lock;/* lock for dev_cmd operations */ struct mutex config_lock;/* lock for configuration operations */ spinlock_t adminq_lock;/* lock for adminq operations */ +refcount_t adminq_refcnt;struct pds_core_dev_info_regs __iomem *info_regs;struct pds_core_dev_cmd_regs __iomem *cmd_regs;struct pds_core_intr __iomem *intr_ctrl;
Signature · 1 line
2.17.1
Brett Creeley
maintainerPATCH 5/6 pds_core: Clear BARs on reset
In reply to Brett CreeleyDuring reset the BARs might be accessed when they are unmapped. This can cause unexpected issues, so fix it by clearing the cached BAR values so they are not accessed until they are re-mapped.
Also, make sure any places that can access the BARs when they are NULL are prevented.
--- drivers/net/ethernet/amd/pds_core/adminq.c | 28 +++++++++++++++------ drivers/net/ethernet/amd/pds_core/core.c | 8 +++++- drivers/net/ethernet/amd/pds_core/dev.c | 9 ++++++- drivers/net/ethernet/amd/pds_core/devlink.c | 3 ++- drivers/net/ethernet/amd/pds_core/fw.c | 3 +++ drivers/net/ethernet/amd/pds_core/main.c | 5 ++++ 6 files changed, 45 insertions(+), 11 deletions(-)
drivers/net/ethernet/amd/pds_core/adminq.c +20 -8
@@ -191,10 +191,16 @@ static int __pdsc_adminq_post(struct pdsc *pdsc,/* Check that the FW is running */ if (!pdsc_is_fw_running (pdsc)) { -u8 fw_status =ioread8 (&pdsc->info_regs->fw_status); - -dev_info (pdsc->dev,"%s: post failed - fw not running %#02x:\n" , - __func__, fw_status); +if (pdsc->info_regs) { +u8 fw_status = +ioread8 (&pdsc->info_regs->fw_status); + +dev_info (pdsc->dev,"%s: post failed - fw not running %#02x:\n" , + __func__, fw_status); + }else { +dev_info (pdsc->dev,"%s: post failed - BARs not setup\n" , + __func__); + } ret = -ENXIO ; goto err_out_unlock;
@@ -266,10 +272,16 @@ int pdsc_adminq_post(struct pdsc *pdsc,break ;if (!pdsc_is_fw_running (pdsc)) { -u8 fw_status =ioread8 (&pdsc->info_regs->fw_status); - -dev_dbg (pdsc->dev,"%s: post wait failed - fw not running %#02x:\n" , - __func__, fw_status); +if (pdsc->info_regs) { +u8 fw_status = +ioread8 (&pdsc->info_regs->fw_status); + +dev_dbg (pdsc->dev,"%s: post wait failed - fw not running %#02x:\n" , + __func__, fw_status); + }else { +dev_dbg (pdsc->dev,"%s: post wait failed - BARs not setup\n" , + __func__); + } err = -ENXIO ;break ; }
drivers/net/ethernet/amd/pds_core/core.c +7 -1
@@ -600,7 +600,13 @@ void pdsc_fw_up(struct pdsc *pdsc)static void pdsc_check_pci_health (struct pdsc *pdsc) { -u8 fw_status =ioread8 (&pdsc->info_regs->fw_status); +u8 fw_status; + +/* some sort of teardown already in progress */ +if (!pdsc->info_regs) +return ; + + fw_status =ioread8 (&pdsc->info_regs->fw_status);/* is PCI broken? */ if (fw_status !=PDS_RC_BAD_PCI )
drivers/net/ethernet/amd/pds_core/dev.c +8 -1
@@ -57,6 +57,9 @@ int pdsc_err_to_errno(enum pds_core_status_code code)bool pdsc_is_fw_running (struct pdsc *pdsc) { +if (!pdsc->info_regs) +return false; + pdsc->fw_status =ioread8 (&pdsc->info_regs->fw_status); pdsc->last_fw_time = jiffies; pdsc->last_hb =ioread32 (&pdsc->info_regs->fw_heartbeat);
@@ -182,13 +185,17 @@ int pdsc_devcmd_locked(struct pdsc *pdsc, union pds_core_dev_cmd *cmd, {int err; +if (!pdsc->cmd_regs) +return -ENXIO ; +memcpy_toio (&pdsc->cmd_regs->cmd, cmd,sizeof (*cmd));pdsc_devcmd_dbell (pdsc); err =pdsc_devcmd_wait (pdsc, cmd->opcode, max_seconds); -memcpy_fromio (comp, &pdsc->cmd_regs->comp,sizeof (*comp));if ((err == -ENXIO || err == -ETIMEDOUT ) && pdsc->wq)queue_work (pdsc->wq, &pdsc->health_work); +else +memcpy_fromio (comp, &pdsc->cmd_regs->comp,sizeof (*comp));return err; }
drivers/net/ethernet/amd/pds_core/devlink.c +2 -1
@@ -111,7 +111,8 @@ int pdsc_dl_info_get(struct devlink *dl, struct devlink_info_req *req,mutex_lock (&pdsc->devcmd_lock); err =pdsc_devcmd_locked (pdsc, &cmd, &comp, pdsc->devcmd_timeout *2 ); -memcpy_fromio (&fw_list, pdsc->cmd_regs->data,sizeof (fw_list)); +if (!err) +memcpy_fromio (&fw_list, pdsc->cmd_regs->data,sizeof (fw_list));mutex_unlock (&pdsc->devcmd_lock);if (err && err != -EIO )return err;
drivers/net/ethernet/amd/pds_core/fw.c +3 -0
@@ -107,6 +107,9 @@ int pdsc_firmware_update(struct pdsc *pdsc, const struct firmware *fw,dev_info (pdsc->dev,"Installing firmware\n" ); +if (!pdsc->cmd_regs) +return -ENXIO ; + dl =priv_to_devlink (pdsc);devlink_flash_update_status_notify (dl,"Preparing to flash" ,NULL ,0 ,0 );
drivers/net/ethernet/amd/pds_core/main.c +5 -0
@@ -37,6 +37,11 @@ static void pdsc_unmap_bars(struct pdsc *pdsc)struct pdsc_dev_bar *bars = pdsc->bars;unsigned int i; + pdsc->info_regs =NULL ; + pdsc->cmd_regs =NULL ; + pdsc->intr_status =NULL ; + pdsc->intr_ctrl =NULL ; +for (i =0 ; i <PDS_CORE_BARS_MAX ; i++) {if (bars[i].vaddr)pci_iounmap (pdsc->pdev, bars[i].vaddr);
Signature · 1 line
2.17.1
Brett Creeley
maintainerPATCH 6/6 pds_core: Rework teardown/setup flow to be more common
In reply to Brett CreeleyCurrently the teardown/setup flow for driver probe/remove is quite a bit different from the reset flows in pdsc_fw_down()/pdsc_fw_up(). One key piece that's missing are the calls to pci_alloc_irq_vectors() and pci_free_irq_vectors(). The pcie reset case is calling pci_free_irq_vectors() on reset_prepare, but not calling the corresponding pci_alloc_irq_vectors() on reset_done. This is causing unexpected/unwanted interrupt behavior due to the adminq interrupt being accidentally put into legacy interrupt mode. Also, the pci_alloc_irq_vectors()/pci_free_irq_vectors() functions are being called directly in probe/remove respectively.
Fix this inconsistency by making the following changes:
1. Always call pdsc_dev_init() in pdsc_setup(), which calls pci_alloc_irq_vectors() and get rid of the now unused pds_dev_reinit().
2. Always free/clear the pdsc->intr_info in pdsc_teardown() since this structure will get re-alloced in pdsc_setup().
3. Move the calls of pci_free_irq_vectors() to pdsc_teardown() since pci_alloc_irq_vectors() will always be called in pdsc_setup()->pdsc_dev_init() for both the probe/remove and reset flows.
4. Make sure to only create the debugfs "identity" entry when it doesn't already exist, which it will in the reset case because it's already been created in the initial call to pdsc_dev_init().
--- drivers/net/ethernet/amd/pds_core/core.c | 13 +++++-------- drivers/net/ethernet/amd/pds_core/core.h | 1 - drivers/net/ethernet/amd/pds_core/debugfs.c | 4 ++++ drivers/net/ethernet/amd/pds_core/dev.c | 7 ------- drivers/net/ethernet/amd/pds_core/main.c | 2 -- 5 files changed, 9 insertions(+), 18 deletions(-)
drivers/net/ethernet/amd/pds_core/core.c +5 -8
@@ -404,10 +404,7 @@ int pdsc_setup(struct pdsc *pdsc, bool init)int numdescs;int err; -if (init) - err =pdsc_dev_init (pdsc); -else - err =pdsc_dev_reinit (pdsc); + err =pdsc_dev_init (pdsc);if (err)return err;
@@ -479,10 +476,9 @@ void pdsc_teardown(struct pdsc *pdsc, bool removing)for (i =0 ; i < pdsc->nintrs; i++)pdsc_intr_free (pdsc, i); -if (removing) { -kfree (pdsc->intr_info); - pdsc->intr_info =NULL ; - } +kfree (pdsc->intr_info); + pdsc->intr_info =NULL ; + pdsc->nintrs =0 ; }if (pdsc->kern_dbpage) {
@@ -490,6 +486,7 @@ void pdsc_teardown(struct pdsc *pdsc, bool removing) pdsc->kern_dbpage =NULL ; } +pci_free_irq_vectors (pdsc->pdev);set_bit (PDSC_S_FW_DEAD , &pdsc->state); }
drivers/net/ethernet/amd/pds_core/core.h +0 -1
@@ -281,7 +281,6 @@ int pdsc_devcmd_locked(struct pdsc *pdsc, union pds_core_dev_cmd *cmd, union pds_core_dev_comp *comp, int max_seconds);int pdsc_devcmd_init (struct pdsc *pdsc);int pdsc_devcmd_reset (struct pdsc *pdsc); -int pdsc_dev_reinit (struct pdsc *pdsc);int pdsc_dev_init (struct pdsc *pdsc);void pdsc_reset_prepare (struct pci_dev *pdev);
drivers/net/ethernet/amd/pds_core/debugfs.c +4 -0
@@ -64,6 +64,10 @@ DEFINE_SHOW_ATTRIBUTE(identity);void pdsc_debugfs_add_ident (struct pdsc *pdsc) { +/* This file will already exist in the reset flow */ +if (debugfs_lookup ("identity" , pdsc->dentry)) +return ; +debugfs_create_file ("identity" ,0400 , pdsc->dentry, pdsc, &identity_fops); }
drivers/net/ethernet/amd/pds_core/dev.c +0 -7
@@ -316,13 +316,6 @@ static int pdsc_identify(struct pdsc *pdsc)return 0 ; } -int pdsc_dev_reinit (struct pdsc *pdsc) -{ -pdsc_init_devinfo (pdsc); - -return pdsc_identify (pdsc); -} -int pdsc_dev_init (struct pdsc *pdsc) {unsigned int nintrs;
drivers/net/ethernet/amd/pds_core/main.c +0 -2
@@ -438,7 +438,6 @@ static void pdsc_remove(struct pci_dev *pdev)mutex_destroy (&pdsc->config_lock);mutex_destroy (&pdsc->devcmd_lock); -pci_free_irq_vectors (pdev);pdsc_unmap_bars (pdsc);pci_release_regions (pdev); }
@@ -470,7 +469,6 @@ void pdsc_reset_prepare(struct pci_dev *pdev)pdsc_stop_health_thread (pdsc);pdsc_fw_down (pdsc); -pci_free_irq_vectors (pdev);pdsc_unmap_bars (pdsc);pci_release_regions (pdev);pci_disable_device (pdev);
Signature · 1 line
2.17.1
This series was applied to netdev/net.git (main) by Jakub Kicinski <[email protected]>:
Hello:
This series was applied to netdev/net.git (main)
by Jakub Kicinski <[email protected]>:
Quoted 10 lines · On Mon, 29 Jan 2024 15:40:29 -0800 you wrote:
This series includes the following changes:
There can be many users of the pds_core's adminq. This includes pds_core's uses and any clients that depend on it. When the
pds_coredevice goes through a reset for any reason the adminq is freed and reconfigured. There are some gaps in the current implementation that will cause crashes during reset if any of the previously mentioned users of the adminq attempt to use it after it's been freed.[...]
Here is the summary with links:
- [net,1/6] pds_core: Prevent health thread from running during reset/remove
https://git.kernel.org/netdev/net/c/d9407ff11809 - [net,2/6] pds_core: Cancel AQ work on teardown https://git.kernel.org/netdev/net/c/d321067e2cfa - [net,3/6] pds_core: Use struct pdsc for the pdsc_adminq_isr private data https://git.kernel.org/netdev/net/c/951705151e50 - [net,4/6] pds_core: Prevent race issues involving the adminq https://git.kernel.org/netdev/net/c/7e82a8745b95 - [net,5/6] pds_core: Clear BARs on reset https://git.kernel.org/netdev/net/c/e96094c1d11c - [net,6/6] pds_core: Rework teardown/setup flow to be more common https://git.kernel.org/netdev/net/c/bc90fbe0c318
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html