iMX/i2som-imx-linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

MLK-21252 drm/bridge: sec-dsim: resolve OVERFLOW_BEFORE_WIDEN warns

Browse Source 
Coverity reports the 'OVERFLOW_BEFORE_WIDEN' warns which can
cause potential expression overflow for multiplying two 32bit
integers and assign the result to a 64bit integer, since no
automatic integer type promotion exists before the multiply
operation, so requires explicit type casting for either one
of operands.

Signed-off-by: Fancy Fang <chen.fang@nxp.com>
(cherry picked from commit 9f10b3172107df0c31031e3fb02d32b6f26696f2)
This commit is contained in:
Fancy Fang
2019-03-30 22:47:54 +08:00
committed by Dong Aisheng
parent 2a0fed9fbb
commit a940cee4fd
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
drivers/gpu/drm/bridge/sec-dsim.c
View File

@ -1193,10 +1193,10 @@ struct dsim_pll_pms *sec_mipi_dsim_calc_pmsk(struct sec_mipi_dsim *dsim)
* Fvco: [1050MHz ~ 2100MHz] (Fvco = ((m + k / 65536) * Fin) / p)
* So, m = Fvco * p / Fin and Fvco > Fin;
*/
pfvco = fvco_range->min * prange->min;
pfvco = (uint64_t)fvco_range->min * prange->min;
mrange->min = max_t(uint32_t, mrange->min,
DIV_ROUND_UP_ULL(pfvco, fin));
pfvco = fvco_range->max * prange->max;
pfvco = (uint64_t)fvco_range->max * prange->max;
mrange->max = min_t(uint32_t, mrange->max,
DIV_ROUND_UP_ULL(pfvco, fin));
@ -1209,7 +1209,7 @@ struct dsim_pll_pms *sec_mipi_dsim_calc_pmsk(struct sec_mipi_dsim *dsim)
/* first determine 'm', then can determine 'p', last determine 's' */
for (m = mrange->min; m <= mrange->max; m++) {
/* p = m * Fin / Fvco */
mfin = m * fin;
mfin = (uint64_t)m * fin;
pr_new.min = max_t(uint32_t, prange->min,
DIV_ROUND_UP_ULL(mfin, fvco_range->max));
pr_new.max = min_t(uint32_t, prange->max,
@ -1220,7 +1220,7 @@ struct dsim_pll_pms *sec_mipi_dsim_calc_pmsk(struct sec_mipi_dsim *dsim)
for (p = pr_new.min; p <= pr_new.max; p++) {
/* s = order_pow_of_two((m * Fin) / (p * Fout)) */
pfout = p * fout;
pfout = (uint64_t)p * fout;
raw_s = DIV_ROUND_CLOSEST_ULL(mfin, pfout);
s_pow_2 = rounddown_pow_of_two(raw_s);