git: 6387c1135921 - main - ath10k: fix a schedule() call
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Tue, 10 Feb 2026 00:46:48 UTC
The branch main has been updated by bz:
URL: https://cgit.FreeBSD.org/src/commit/?id=6387c1135921fbc86ab0f3c1508a3719ca9cac54
commit 6387c1135921fbc86ab0f3c1508a3719ca9cac54
Author: Bjoern A. Zeeb <bz@FreeBSD.org>
AuthorDate: 2025-12-08 00:00:05 +0000
Commit: Bjoern A. Zeeb <bz@FreeBSD.org>
CommitDate: 2026-02-09 21:49:45 +0000
ath10k: fix a schedule() call
Using kern_yield does not seem ideal either for polling here; use a
schedule_timeout() call from LinuxKPI instead and use about 100
iterations per second.
Improve an error message to include the error code so we have a better
idea of what happens.
MFC after: 3 days
---
sys/contrib/dev/athk/ath10k/bmi.c | 5 +++++
sys/contrib/dev/athk/ath10k/pci.c | 7 +++++--
2 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/sys/contrib/dev/athk/ath10k/bmi.c b/sys/contrib/dev/athk/ath10k/bmi.c
index 6dd92d29f631..30dded5b6bbb 100644
--- a/sys/contrib/dev/athk/ath10k/bmi.c
+++ b/sys/contrib/dev/athk/ath10k/bmi.c
@@ -66,7 +66,12 @@ int ath10k_bmi_get_target_info(struct ath10k *ar,
ret = ath10k_hif_exchange_bmi_msg(ar, &cmd, cmdlen, &resp, &resplen);
if (ret) {
+#if defined(__linux__)
ath10k_warn(ar, "unable to get target info from device\n");
+#elif defined(__FreeBSD__)
+ ath10k_warn(ar, "unable to get target info from device: %d\n",
+ ret);
+#endif
return ret;
}
diff --git a/sys/contrib/dev/athk/ath10k/pci.c b/sys/contrib/dev/athk/ath10k/pci.c
index e613eaf24797..e9b8573a698b 100644
--- a/sys/contrib/dev/athk/ath10k/pci.c
+++ b/sys/contrib/dev/athk/ath10k/pci.c
@@ -2340,8 +2340,11 @@ static int ath10k_pci_bmi_wait(struct ath10k *ar,
#if defined(__linux__)
schedule();
#elif defined(__FreeBSD__)
- /* Using LinuxKPI we'll hang for-ever as there's no wake_up */
- kern_yield(PRI_USER);
+ /*
+ * Using LinuxKPI's schedule() will hang for-ever as there is
+ * no wake_up. Poll about 100 times per second until timeout.
+ */
+ schedule_timeout(BMI_COMMUNICATION_TIMEOUT_HZ/300);
#endif
}