possible typo in ancient radeon code
Hello FB devs,
This ancient code from 1da177e4c3f4: "Linux-2.6.12-rc2" from Apr 16, 2005, leads to the following static checker warning:
drivers/video/fbdev/aty/radeon_pm.c:417 radeon_pm_enable_dynamic_mode()
warn: we tested 'tmp & (1 << 21)' before and it was 'true'
include/video/radeon.h
1216 #define MCLK_CNTL__MRDCKA1_SOUTSEL_MASK 0x0c000000L 1217 #define MCLK_CNTL__MRDCKB0_SOUTSEL_MASK 0x30000000L 1218 #define MCLK_CNTL__MRDCKB1_SOUTSEL_MASK 0xc0000000L 1219 #define MCLK_CNTL__R300_DISABLE_MC_MCLKA (1 << 21) 1220 #define MCLK_CNTL__R300_DISABLE_MC_MCLKB (1 << 21)
These are probably supposed to refer to different bits.
1221
1222 // MCLK_MISC
drivers/video/fbdev/aty/radeon_pm.c
411 /* Some releases of vbios have set DISABLE_MC_MCLKA
412 * and DISABLE_MC_MCLKB bits in the vbios table. Setting these 413 * bits will cause H/W hang when reading video memory with dynamic 414 * clocking enabled. 415 */
416 if ((tmp & MCLK_CNTL__R300_DISABLE_MC_MCLKA) &&
417 (tmp & MCLK_CNTL__R300_DISABLE_MC_MCLKB)) {
The duplicate check here causes a warning. I guess it probably doesn't matter if no one complains...
418 /* If both bits are set, then check the active channels */
419 tmp = INPLL(pllMCLK_CNTL);
420 if (rinfo->vram_width = 64) { 421 if (INREG(MEM_CNTL) & R300_MEM_USE_CD_CH_ONLY)
422 tmp &= ~MCLK_CNTL__R300_DISABLE_MC_MCLKB;
423 else
424 tmp &= ~MCLK_CNTL__R300_DISABLE_MC_MCLKA;
425 } else {
426 tmp &= ~(MCLK_CNTL__R300_DISABLE_MC_MCLKA |
427 MCLK_CNTL__R300_DISABLE_MC_MCLKB);
428 }
429 }
regards,
dan carpenter
CC BenH
Quoted 4 lines · Dan Carpenter
Hello FB devs,
This ancient code from
1da177e4c3f4: "Linux-2.6.12-rc2" from Apr 16, 2005, leads to the following static checker warning:
Nah, it's a little bit older (in full-history-linux ;-)
commit 9ff343f3d155aa35f79bae8607d7c4500d7ef848
Author: Benjamin Herrenschmidt <[email protected]> Date: Thu Feb 10 16:03:30 2005 -0800
[PATCH] radeonfb update
Quoted 28 lines
drivers/video/fbdev/aty/radeon_pm.c:417radeon_pm_enable_dynamic_mode()
warn: we tested 'tmp & (1 << 21)' before and it was 'true'
include/video/radeon.h1216 #define MCLK_CNTL__MRDCKA1_SOUTSEL_MASK 0x0c000000L 1217 #define MCLK_CNTL__MRDCKB0_SOUTSEL_MASK 0x30000000L 1218 #define MCLK_CNTL__MRDCKB1_SOUTSEL_MASK 0xc0000000L 1219 #define MCLK_CNTL__R300_DISABLE_MC_MCLKA (1 << 21) 1220 #define MCLK_CNTL__R300_DISABLE_MC_MCLKB (1 << 21)These are probably supposed to refer to different bits.
1221
1222 //MCLK_MISC
drivers/video/fbdev/aty/radeon_pm.c
411 /* Some releases of vbios have setDISABLE_MC_MCLKA412 * and DISABLE_MC_MCLKB bits in the vbios table. Setting these 413 * bits will cause H/W hang when reading video memory with dynamic 414 * clocking enabled. 415 */416 if ((tmp &
MCLK_CNTL__R300_DISABLE_MC_MCLKA) &&
417 (tmp &MCLK_CNTL__R300_DISABLE_MC_MCLKB)) {The duplicate check here causes a warning. I guess it probably doesn't matter if no one complains...
Quoted 12 lines
418 /* If both bits are set, then check the active channels */
419 tmp = INPLL(pllMCLK_CNTL);
420 if (rinfo->vram_width = 64) {
421 if (INREG(MEM_CNTL) & R300_MEM_USE_CD_CH_ONLY)
422 tmp &= ~MCLK_CNTL__R300_DISABLE_MC_MCLKB;
423 else
424 tmp &= ~MCLK_CNTL__R300_DISABLE_MC_MCLKA;
425 } else {
426 tmp &= ~(MCLK_CNTL__R300_DISABLE_MC_MCLKA |
427 MCLK_CNTL__R300_DISABLE_MC_MCLKB);
428 }
429 }Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds