On Wed, Nov 05, 2014 at 08:09:46AM +0100, Takashi Iwai wrote:
At Wed, 5 Nov 2014 12:31:50 +0530, Sudip Mukherjee wrote:
On Tue, Nov 04, 2014 at 10:28:01AM -0800, Joe Perches wrote:
On Tue, 2014-11-04 at 23:47 +0530, Sudip Mukherjee wrote:
commit "b5b4a41b392960010fccf1f9ccf8334d612bd450" was dereferencing chip after it has been freed. This patch fixes that and at the same time removes some debugging messages, which are unnecessary, as they are just printing information about entry and exit from a function. This information we can easily get from ftrace.
Mostly true.
diff --git a/sound/pci/echoaudio/echoaudio.c b/sound/pci/echoaudio/echoaudio.c
[]
@@ -763,10 +755,8 @@ static int pcm_trigger(struct snd_pcm_substream *substream, int cmd) spin_lock(&chip->lock); switch (cmd) { case SNDRV_PCM_TRIGGER_RESUME:
case SNDRV_PCM_TRIGGER_START: case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:dev_dbg(chip->card->dev, "pcm_trigger resume\n");
for (i = 0; i < DSP_MAXPIPES; i++) { if (channelmask & (1 << i)) { pipe = chip->substream[i]->runtime->private_data;dev_dbg(chip->card->dev, "pcm_trigger start\n");
These may not be useful but this switch state information can't be got from ftrace.
is this information required? trigger callback will be called everytime pcm is started, stopped or paused. then log will be full of just these messages. if it is required then i can send a v2, if not required then also i can send another cleanup patch to clear the other prints in the switch-case of other devices in echoaudio.
There has been never such "requirement". It's merely a debug code.
then i will send you a v2 of this patch which will also cleanup similar debug messages in the switch-case of other devices in echoaudio.
thanks sudip
Takashi