Re: [ 031/100] iwlwifi: always copy first 16 bytes of commands

[Ben Hutchings]

On Thu, 2013-03-14 at 01:50 +0000, Ben Hutchings wrote:
> On Tue, 2013-03-12 at 15:31 -0700, Greg Kroah-Hartman wrote:
> > 3.8-stable review patch.  If anyone has any objections, please let me know.
> > 
> > ------------------
> > 
> > From: Johannes Berg <johannes.berg@xxxxxxxxx>
> > 
> > commit 8a964f44e01ad3bbc208c3e80d931ba91b9ea786 upstream.
> [...]
> > --- a/drivers/net/wireless/iwlwifi/pcie/tx.c
> > +++ b/drivers/net/wireless/iwlwifi/pcie/tx.c
> [...]
> > @@ -1221,14 +1238,31 @@ static int iwl_pcie_enqueue_hcmd(struct
> >  
> >  	/* and copy the data that needs to be copied */
> >  	cmd_pos = offsetof(struct iwl_device_cmd, payload);
> > +	copy_size = sizeof(out_cmd->hdr);
> >  	for (i = 0; i < IWL_MAX_CMD_TFDS; i++) {
> > -		if (!cmd->len[i])
> > +		int copy = 0;
> > +
> > +		if (!cmd->len)
> >  			continue;
> 
> cmd->len is an array, so the new condition is always false.  Shouldn't
> it be 'if (!cmdlen[i])'?

To answer myself: no, it should still be 'if (!cmd->len[i])' as this
loop needs to include input fragments that will be completely copied
into the header fragment.

Ben.

> Ben.
> 
> > -		if (cmd->dataflags[i] & (IWL_HCMD_DFL_NOCOPY |
> > -					 IWL_HCMD_DFL_DUP))
> > -			break;
> > -		memcpy((u8 *)out_cmd + cmd_pos, cmd->data[i], cmd->len[i]);
> > -		cmd_pos += cmd->len[i];
> > +
> > +		/* need at least IWL_HCMD_MIN_COPY_SIZE copied */
> > +		if (copy_size < IWL_HCMD_MIN_COPY_SIZE) {
> > +			copy = IWL_HCMD_MIN_COPY_SIZE - copy_size;
> > +
> > +			if (copy > cmd->len[i])
> > +				copy = cmd->len[i];
> > +		}
> > +
> > +		/* copy everything if not nocopy/dup */
> > +		if (!(cmd->dataflags[i] & (IWL_HCMD_DFL_NOCOPY |
> > +					   IWL_HCMD_DFL_DUP)))
> > +			copy = cmd->len[i];
> > +
> > +		if (copy) {
> > +			memcpy((u8 *)out_cmd + cmd_pos, cmd->data[i], copy);
> > +			cmd_pos += copy;
> > +			copy_size += copy;
> > +		}
> >  	}
> >  
> >  	WARN_ON_ONCE(txq->entries[idx].copy_cmd);
> [...]
> 

-- 
Ben Hutchings
Humans are not rational beings; they are rationalising beings.

Attachment: signature.asc
Description: This is a digitally signed message part