The driver appears to work reliably (i.e. lockup-free) with a card based on this chip. However, I have not explored whether we could get anything in terms of 3D performance "for free" just by setting some magic bits in those registers that are still a mystery to us.tags/mesa_20050715
| @@ -211,6 +211,7 @@ static const struct dri_debug_control debug_control[] = { | |||
| #define PCI_CHIP_R300_NF 0x4E46 | |||
| #define PCI_CHIP_R300_NG 0x4E47 | |||
| #define PCI_CHIP_R350_NH 0x4E48 | |||
| #define PCI_CHIP_R420_JI 0x4A49 | |||
| #define PCI_CHIP_R420_JK 0x4a4b | |||
| #endif | |||
| @@ -393,6 +394,7 @@ static radeonScreenPtr radeonCreateScreen(__DRIscreenPrivate * sPriv) | |||
| screen->chipset = RADEON_CHIP_R300; | |||
| break; | |||
| case PCI_CHIP_R420_JI: | |||
| case PCI_CHIP_R420_JK: | |||
| screen->chipset = RADEON_CHIP_R420; | |||
| break; | |||