topic Here Adriaan van Os, in Intel® Integrated Performance Primitives

ippiFilterBorder bug

Adriaan_van_Os — Wed, 11 Mar 2020 11:23:44 GMT

I noticed a bug in ippiFilterBorder where ippiFilterBorder_16u_C4R and ippiFilterBorder_16u_C1R (and rarely also ippiFilterBorder_32f_C4R and ippiFilterBorder_32f_C1R, but never ippiFilterBorder_8u_C4R or ippiFilterBorder_8u_C1R) write random pixels in parts of the last row of the target image when the convolution kernel has a width of 1 and a height > 1. The result seems to depend on the border type, but does happen with all border types.

The workaround is to apply a horizontal kernel (with height 1 and width > 1) to a transposed image and then to transpose the result back.

This happens with IPP version 9.0, but I can't reproduce it with version 2018-2 or 2019-0. However, I don't see an entry for this in the IPP bug fix logs. Can you confirm that this is something that has been fixed between version 9.0 and 2018-2 ?

Regards,

Adriaan van Os

Here Adriaan van Os,

Ruqiu_C_Intel — Mon, 16 Mar 2020 02:20:33 GMT

Here Adriaan van Os,

Thanks a lot for raising the issue in old library IPP 9.0 legacy. We will double check the fixed bug logs internally.

BRs,

Ruqiu

But can you confirm that this

Adriaan_van_Os — Mon, 16 Mar 2020 03:49:02 GMT

But can you confirm that this is something that has been fixed between version 9.0 and 2018-2 ?

Regards,

Adriaan van Os

Hello Adriaan van Os,

Ruqiu_C_Intel — Wed, 18 Mar 2020 13:18:04 GMT

Hello Adriaan van Os,

We did fix some ippiFilterBordersome bugs since version 9.0.

Best Regards,

Ruqiu