On Wed, 2023-06-14 at 12:03 +0200, Johannes Berg wrote:
> On Mon, 2023-06-12 at 18:51 +0300, gregory.greenman@xxxxxxxxx wrote:
> > From: Avraham Stern <avraham.stern@xxxxxxxxx>
> >
> > Add an entry for setting the maximum TXOP time in microseconds.
> > The configured value can be read from the same entry.
> >
> > Signed-off-by: Avraham Stern <avraham.stern@xxxxxxxxx>
> > Signed-off-by: Gregory Greenman <gregory.greenman@xxxxxxxxx>
> > ---
> > .../wireless/intel/iwlwifi/mvm/debugfs-vif.c | 38 +++++++++++++++++++
> > drivers/net/wireless/intel/iwlwifi/mvm/mvm.h | 3 ++
> > .../net/wireless/intel/iwlwifi/mvm/rs-fw.c | 3 ++
> > 3 files changed, 44 insertions(+)
> >
> > diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c b/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c
> > index 3613b1fdc5d9..762bce70e262 100644
> > --- a/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c
> > +++ b/drivers/net/wireless/intel/iwlwifi/mvm/debugfs-vif.c
> > @@ -679,6 +679,42 @@ static ssize_t iwl_dbgfs_quota_min_read(struct file *file,
> > return simple_read_from_buffer(user_buf, count, ppos, buf, len);
> > }
> >
> > +static ssize_t iwl_dbgfs_max_tx_op_write(struct ieee80211_vif *vif, char *buf,
> > + size_t count, loff_t *ppos)
> > +{
> > + struct iwl_mvm_vif *mvmvif = iwl_mvm_vif_from_mac80211(vif);
> > + struct iwl_mvm *mvm = mvmvif->mvm;
> > + u16 value;
> > + int ret;
> > +
> > + ret = kstrtou16(buf, 0, &value);
> > + if (ret)
> > + return ret;
> > +
> > + mutex_lock(&mvm->mutex);
> > + mvmvif->max_tx_op = value;
> > + mutex_unlock(&mvm->mutex);
>
> This is a strange debugfs file, it only takes effect the next time
> iwl_mvm_rs_fw_rate_init() is called, which can be kind of random? Seems
> like it should force a reinit?
>
> johannes
You're right. We need to add some fixes to this commit.
Let's drop it meanwhile, I'll send a new version later.
I've verified that there're no any conflicts without this patch.
