Good day,

On Sat, Oct 23, 2021 at 10:09:14PM +0200, Christophe JAILLET wrote:

Le 23/10/2021 à 21:36, Joe Perches a écrit :

...

On Sat, 2021-10-23 at 21:24 +0200, Christophe JAILLET wrote:

...

'drvdata->chs.guaranteed' is a bitmap. So use 'devm_bitmap_kzalloc()' to simplify code, improve the semantic and avoid some open-coded arithmetic in allocator arguments.

[]

...

diff --git a/drivers/hwtracing/coresight/coresight-stm.c b/drivers/hwtracing/coresight/coresight-stm.c

[]

...

@@ -862,7 +862,6 @@ static int stm_probe(struct amba_device *adev, const struct amba_id *id) struct stm_drvdata *drvdata; struct resource *res = &adev->res; struct resource ch_res;

• size_t bitmap_size; struct coresight_desc desc = { 0 }; desc.name = coresight_alloc_device_name(&stm_devs, dev);

@@ -904,9 +903,7 @@ static int stm_probe(struct amba_device *adev, const struct amba_id *id) else drvdata->numsp = stm_num_stimulus_port(drvdata);

• bitmap_size = BITS_TO_LONGS(drvdata->numsp) * sizeof(long);
• guaranteed = devm_kzalloc(dev, bitmap_size, GFP_KERNEL);

• guaranteed = devm_bitmap_zalloc(dev, drvdata->numsp, GFP_KERNEL); if (!guaranteed) return -ENOMEM; drvdata->chs.guaranteed = guaranteed;

guaranteed is also pretty useless

I agree, but removing it would make the line with devm_bitmap_zalloc() 86 chars. This would not be consistent with the rest of the file and would (IMHO) require splitting.

Let see if the maintainer prefer saving one additional line of code, or keeping the logic in place.

I think we can get rid of @guaranteed and splitting is fine with me:

drvdata->chs.guaranteed = devm_bitmap_zalloc(dev, drvdata->numsp, GFP_KERNEL);

Thanks, Mathieu

CJ