Hello Benoit Parrot,
The patch 343e89a792a5: "[media] media: ti-vpe: Add CAL v4l2 camera
capture driver" from Jan 6, 2016, leads to the following static
checker warning:
drivers/media/platform/ti-vpe/cal.c:1218 cal_enum_frameintervals()
info: ignoring unreachable code.
drivers/media/platform/ti-vpe/cal.c
1197 /* timeperframe is arbitrary and continuous */
1198 static int cal_enum_frameintervals(struct file *file, void *priv,
1199 struct v4l2_frmivalenum *fival)
1200 {
1201 struct cal_ctx *ctx = video_drvdata(file);
1202 const struct cal_fmt *fmt;
1203 struct v4l2_subdev_frame_size_enum fse;
1204 int ret;
1205
1206 if (fival->index)
1207 return -EINVAL;
1208
1209 fmt = find_format_by_pix(ctx, fival->pixel_format);
1210 if (!fmt)
1211 return -EINVAL;
1212
1213 /* check for valid width/height */
1214 ret = 0;
1215 fse.pad = 0;
1216 fse.code = fmt->code;
1217 fse.which = V4L2_SUBDEV_FORMAT_ACTIVE;
1218 for (fse.index = 0; ; fse.index++) {
^^^^^^^^^^^
We never do this increment. Is this stub code that was accidentally
committed or how is this loop supposed to work?
1219 ret = v4l2_subdev_call(ctx->sensor, pad, enum_frame_size,
1220 NULL, &fse);
1221 if (ret)
1222 return -EINVAL;
1223
1224 if ((fival->width == fse.max_width) &&
1225 (fival->height == fse.max_height))
1226 break;
1227 else if ((fival->width >= fse.min_width) &&
1228 (fival->width <= fse.max_width) &&
1229 (fival->height >= fse.min_height) &&
1230 (fival->height <= fse.max_height))
1231 break;
1232
1233 return -EINVAL;
1234 }
1235
1236 fival->type = V4L2_FRMIVAL_TYPE_DISCRETE;
1237 fival->discrete.numerator = 1;
1238 fival->discrete.denominator = 30;
1239
1240 return 0;
regards,
dan carpenter
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
