bnxt_en: Update for net-next
Michael Chan
maintainercover bnxt_en: Update for net-next
The first patch changes the FW message timeout threshold for a warning message. The second patch adjusts the ethtool -w coredump length to suppress a warning. The last 2 patches are small cleanup patches for the bnxt_ulp RoCE auxbus code.
Kalesh AP (2):
bnxt_en: Remove unused field "ref_count" in struct bnxt_ulp
bnxt_en: Remove unused macros in bnxt_ulp.h
Michael Chan (1):
bnxt_en: Change FW message timeout warning
Shruti Parab (1):
bnxt_en: Report the ethtool coredump length after copying the coredump
drivers/net/ethernet/broadcom/bnxt/bnxt_coredump.c | 11 +++++++++-- drivers/net/ethernet/broadcom/bnxt/bnxt_hwrm.h | 2 +- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 5 ----- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 4 ---- 4 files changed, 10 insertions(+), 12 deletions(-)
Signature · 1 line
2.30.1
Michael Chan
maintainerPATCH 1/4 bnxt_en: Change FW message timeout warning
The firmware advertises a "hwrm_cmd_max_timeout" value to the driver for NVRAM and coredump related functions that can take tens of seconds to complete. The driver polls for the operation to complete under mutex and may trigger hung task watchdog warning if the wait is too long. To warn the user about this, the driver currently prints a warning if this advertised value exceeds 40 seconds:
Device requests max timeout of %d seconds, may trigger hung task watchdog
The original hope was that newer FW would be able to reduce this timeout below 40 seconds. But 60 seconds is the timeout on most production FW and cannot be reduced further. Change the driver's warning threshold to 60 seconds to avoid triggering this warning on all production devices.
--- drivers/net/ethernet/broadcom/bnxt/bnxt_hwrm.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/broadcom/bnxt/bnxt_hwrm.h +1 -1
@@ -58,7 +58,7 @@ void hwrm_update_token(struct bnxt *bp, u16 seq, enum bnxt_hwrm_wait_state s);#define BNXT_HWRM_MAX_REQ_LEN (bp->hwrm_max_req_len)#define BNXT_HWRM_SHORT_REQ_LEN sizeof(struct hwrm_short_input) -#define HWRM_CMD_MAX_TIMEOUT 40000U +#define HWRM_CMD_MAX_TIMEOUT 60000U#define SHORT_HWRM_CMD_TIMEOUT 20#define HWRM_CMD_TIMEOUT (bp->hwrm_cmd_timeout)#define HWRM_RESET_TIMEOUT ((HWRM_CMD_TIMEOUT) * 4)
Signature · 1 line
2.30.1
Jakub Kicinski
maintainerRe: [PATCH net-next 1/4] bnxt_en: Change FW message timeout warning
Quoted 27 lines · On Tue, 15 Apr 2025 10:48:15 -0700 Michael Chan wrote:
The firmware advertises a "
hwrm_cmd_max_timeout" value to the driver for NVRAM and coredump related functions that can take tens of seconds to complete. The driver polls for the operation to complete under mutex and may trigger hung task watchdog warning if the wait is too long. To warn the user about this, the driver currently prints a warning if this advertised value exceeds 40 seconds:Device requests max timeout of %d seconds, may trigger hung task watchdog
The original hope was that newer FW would be able to reduce this timeout below 40 seconds. But 60 seconds is the timeout on most production FW and cannot be reduced further. Change the driver's warning threshold to 60 seconds to avoid triggering this warning on all production devices.
Reviewed-by Kalesh AP <[email protected]>Reviewed-by Pavan Chebbi <[email protected]>Reviewed-by Andy Gospodarek <[email protected]>Signed-off-by Michael Chan <[email protected]>--- drivers/net/ethernet/broadcom/bnxt/bnxt_hwrm.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)drivers/net/ethernet/broadcom/bnxt/bnxt_hwrm.h +0
-0
drivers/net/ethernet/broadcom/bnxt/bnxt_hwrm.h void hwrm_update_token(struct bnxt *bp, u16 seq, enum bnxt_hwrm_wait_state s); · line 64
Quoted 8 lines
drivers/net/ethernet/broadcom/bnxt/bnxt_hwrm.h +1
-1@@ -58,7 +58,7 @@ void hwrm_update_token(struct bnxt *bp, u16 seq, enum bnxt_hwrm_wait_state s);#define BNXT_HWRM_MAX_REQ_LEN (bp->hwrm_max_req_len)#define BNXT_HWRM_SHORT_REQ_LEN sizeof(struct hwrm_short_input) -#define HWRM_CMD_MAX_TIMEOUT 40000U +#define HWRM_CMD_MAX_TIMEOUT 60000U#define SHORT_HWRM_CMD_TIMEOUT 20#define HWRM_CMD_TIMEOUT (bp->hwrm_cmd_timeout)#define HWRM_RESET_TIMEOUT ((HWRM_CMD_TIMEOUT) * 4)
sysctl_hung_task_timeout_secs is an exported symbol, and it defaults to 120. Should you not use it in the warning (assuming I understand the intent there)?
Michael Chan
maintainerRe: [PATCH net-next 1/4] bnxt_en: Change FW message timeout warning
Quoted 3 lines · Jakub Kicinski
sysctl_hung_task_timeout_secsis an exported symbol, and it defaults to 120. Should you not use it in the warning (assuming I understand the intent there)?
Yes, we have considered that. This is only printed once at driver load time, but the sysctl value can be changed at any time after the driver is loaded. So we just want to use a reasonable value well below the default sysctl_hung_task_timeout_secs value as the threshold.
But we can reference and compare with the sysctl_hung_task_timeout_secs value if that makes more sense.
Jakub Kicinski
maintainerRe: [PATCH net-next 1/4] bnxt_en: Change FW message timeout warning
Quoted 5 lines · On Wed, 16 Apr 2025 14:41:10 -0700 Michael Chan wrote:
Quoted 3 lines · Jakub Kicinski
sysctl_hung_task_timeout_secsis an exported symbol, and it defaults to 120. Should you not use it in the warning (assuming I understand the intent there)?
Quoted 8 lines
Yes, we have considered that. This is only printed once at driver load time, but the sysctl value can be changed at any time after the driver is loaded. So we just want to use a reasonable value well below the default
sysctl_hung_task_timeout_secsvalue as the threshold.But we can reference and compare with the
sysctl_hung_task_timeout_secsvalue if that makes more sense.
I see your point. We could also check againstCONFIG_DEFAULT_HUNG_TASK_TIMEOUT ?
I noticed that some arches set this value really low (10 or 20 sec), it may be worth warning the users in such cases.
Michael Chan
maintainerRe: [PATCH net-next 1/4] bnxt_en: Change FW message timeout warning
Quoted 13 lines · Jakub Kicinski
Quoted 11 lines · On Wed, 16 Apr 2025 14:41:10 -0700 Michael Chan wrote:
Quoted 3 lines · Jakub Kicinski
sysctl_hung_task_timeout_secsis an exported symbol, and it defaults to 120. Should you not use it in the warning (assuming I understand the intent there)?Yes, we have considered that. This is only printed once at driver load time, but the sysctl value can be changed at any time after the driver is loaded. So we just want to use a reasonable value well below the default
sysctl_hung_task_timeout_secsvalue as the threshold.
Quoted 8 lines
Quoted 2 lines
But we can reference and compare with the
sysctl_hung_task_timeout_secsvalue if that makes more sense.I see your point. We could also check against
CONFIG_DEFAULT_HUNG_TASK_TIMEOUT?I noticed that some arches set this value really low (10 or 20 sec), it may be worth warning the users in such cases.
OK, that sounds good. I will check CONFIG_DEFAULT_HUNG_TASK_TIMEOUT in v2. Thanks.
Michael Chan
maintainerPATCH 2/4 bnxt_en: Report the ethtool coredump length after copying the coredump
In reply to Michael ChanFrom: Shruti Parab <[email protected]>
ethtool first calls .get_dump_flags() to get the dump length. For coredump, the driver calls the FW to get the coredump length (L1). The min. of L1 and the user specified length is then passed to .get_dump_data() (L2) to get the coredump. The actual coredump length retrieved by the FW (L3) during .get_dump_data() may be smaller than L1. This length discrepancy will trigger a WARN_ON() in ethtool_get_dump_data().
ethtool has already vzalloc'ed a buffer with size L1. Just report the coredump length as L2 even though the actual coredump length L3 may be smaller. The extra zero padding does not matter. This will prevent the warning that may alarm the user.
For correctness, only do the final length update if there is no error.
--- drivers/net/ethernet/broadcom/bnxt/bnxt_coredump.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt_coredump.c +9 -2
@@ -496,9 +496,16 @@ static int __bnxt_get_coredump(struct bnxt *bp, u16 dump_type, void *buf, start_utc, coredump.total_segs +1 , rc);kfree (coredump.data); - *dump_len +=sizeof (struct bnxt_coredump_record ); -if (rc == -ENOBUFS ) +if (!rc) { + *dump_len +=sizeof (struct bnxt_coredump_record ); +/* The actual coredump length can be smaller than the FW +* reported length earlier. Use the ethtool provided length. +*/ +if (buf_len) + *dump_len = buf_len; + }else if (rc == -ENOBUFS ) {netdev_err (bp->dev,"Firmware returned large coredump buffer\n" ); + }return rc; }
Signature · 1 line
2.30.1
Michael Chan
maintainerPATCH 3/4 bnxt_en: Remove unused field "ref_count" in struct bnxt_ulp
In reply to Michael ChanFrom: Kalesh AP <[email protected]>
The "ref_count" field in struct bnxt_ulp is unused after commit a43c26fa2e6c ("RDMA/bnxt_re: Remove the sriov config callback"). So we can just remove it now.
--- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c | 5 ----- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 1 - 2 files changed, 6 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.c +0 -5
@@ -148,7 +148,6 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev)struct net_device *dev = edev->net;struct bnxt *bp =netdev_priv (dev);struct bnxt_ulp *ulp; -int i =0 ; ulp = edev->ulp_tbl;netdev_lock (dev);
@@ -164,10 +163,6 @@ void bnxt_unregister_dev(struct bnxt_en_dev *edev)synchronize_rcu (); ulp->max_async_event_id =0 ; ulp->async_events_bmap =NULL ; -while (atomic_read (&ulp->ref_count) !=0 && i <10 ) { -msleep (100 ); - i++; - }mutex_unlock (&edev->en_dev_lock);netdev_unlock (dev);return ;
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +0 -1
@@ -50,7 +50,6 @@ struct bnxt_ulp {unsigned long *async_events_bmap;u16 max_async_event_id;u16 msix_requested; -atomic_t ref_count; };struct bnxt_en_dev {
Signature · 1 line
2.30.1
Michael Chan
maintainerPATCH 4/4 bnxt_en: Remove unused macros in bnxt_ulp.h
In reply to Michael ChanFrom: Kalesh AP <[email protected]>
BNXT_ROCE_ULP and BNXT_MAX_ULP are no longer used. Remove them to clean up the code.
--- drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h | 3 --- 1 file changed, 3 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt_ulp.h +0 -3
@@ -10,9 +10,6 @@#ifndef BNXT_ULP_H #define BNXT_ULP_H -#define BNXT_ROCE_ULP 0 -#define BNXT_MAX_ULP 1 -#define BNXT_MIN_ROCE_CP_RINGS 2#define BNXT_MIN_ROCE_STAT_CTXS 1
Signature · 1 line
2.30.1