Lines Matching defs:bits
429 * representation to float is given by \f$\frac{x}{2^{bits} - 1}\f$.
439 * \f$max\left(\frac{x}{2^{bits - 1} - 1}, -1\right)\f$.
449 * As with IEEE floats, the high bits are the exponent and the low bits are
463 * IEEE float16 except that the sign bit and the bottom mantissa bits are
471 * the usual number of mantissa and exponent bits
563 * This differs from the HW enum values used to represent tiling. The bits
748 * On Skylake, this is increased to two bits per cache line pair with 0x0
774 * Starting with Tigerlake, each CCS value is 4 bits per cache line pair in
1172 * In other words, the sample index is swizzled into the low 6 bits of the
1314 uint8_t bits; /**< Size in bits */
1359 * The size (in bits per block) of a single surface element
1384 * The exact value of this field depends heavily on the bits-per-block of
1484 * number of bits per pixel or else memory calculations will go wrong.
2006 return fmtl->channels.r.bits > 0 &&
2007 fmtl->channels.g.bits > 0 &&
2008 fmtl->channels.b.bits > 0 &&
2009 fmtl->channels.a.bits == 0;
2017 return fmtl->channels.r.bits > 0 &&
2018 fmtl->channels.g.bits > 0 &&
2019 fmtl->channels.b.bits > 0 &&
2020 fmtl->channels.a.bits > 0 &&
2368 layout->channels.name.bits == 0)
2783 * @param[in] bpb The size of the surface format in bits per
2824 * @param[in] bpb The size of the surface format in bits per
2947 * @brief Computes masks that may be used to select the bits of the X