search for: nv_ro16

.../drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c +++ b/drivers/gpu/drm/nouveau/core/subdev/bios/rammap.c @@ -84,6 +84,10 @@ nvbios_rammapEp(struct nouveau_bios *bios, int idx, p->rammap_ver = *ver; p->rammap_hdr = *hdr; switch (!!data * *ver) { + case 0x10: + p->rammap_min = nv_ro16(bios, data + 0x00); + p->rammap_max = nv_ro16(bios, data + 0x02); + break; case 0x11: p->rammap_min = nv_ro16(bios, data + 0x00); p->rammap_max = nv_ro16(bios, data + 0x02); -- 1.9.3

Patch 1 fixes nva3 bailing due to not finding the right ramcfg Patch 2 is a resend rebased on 3.17.0-rc4 for setting the vblank period Patch 3-5 handle writes to per-partition registers, for which NVA3 does not have special broadcast regs available. Patch 6 removes local structs from NVA3 reclocking in favour of the already existing "ram->base." variables, like in NVE0 As always,

...for (i = 0; i < bios->size; i += 4) - ((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i); - } + if (!bios->data) + goto out; + + for (i = 0; i < bios->size; i += 4) + ((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i); /* check the PCI record header */ pcir = nv_ro16(bios, 0x0018); -- 1.9.1

...if (!bios->data) >> + goto out; >> + >> + for (i = 0; i < bios->size; i += 4) >> + ((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i); >> >> /* check the PCI record header */ >> pcir = nv_ro16(bios, 0x0018); >> > > This patch never got reviewed or pushed. It should also be sent as a fix for > the next 3.15 rc, if possible. Apologies, this was indeed overlooked. I've merged it. Thanks. > > _______________________________________________ > Nouveau mailing lis...

...((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i); > - } > + if (!bios->data) > + goto out; > + > + for (i = 0; i < bios->size; i += 4) > + ((u32 *)bios->data)[i/4] = nv_rd32(bios, 0x300000 + i); > > /* check the PCI record header */ > pcir = nv_ro16(bios, 0x0018); > This patch never got reviewed or pushed. It should also be sent as a fix for the next 3.15 rc, if possible.

Patch 1 and 2 implement wait-for-vblank, required to remove flicker when reclocking memory Patch 3 and 4 allow me to do things between waiting for VBLANK and disabling FB, like pause PFIFO and wait for the engines to idle. This minimises the time PFIFO is paused, thus maximises performance. The rest of the patches speak for themselves. As the actual memory reclocking script is still somewhat prone

From: Martin Peres <martin.peres at labri.fr> This should fix a deadlock that has been reported to us where fan_update() would hold the fan lock and try to grab the alarm_program_lock to reschedule an update. On an other CPU, the alarm_program_lock would have been taken before calling fan_update(), leading to a deadlock. We should Cc: <stable at vger.kernel.org> # 3.9+ Reported-by:

...; +#include <subdev/bios/fan.h> + +u16 +nvbios_fan_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len) +{ + struct bit_entry bit_P; + u16 fan = 0x0000; + + if (!bit_entry(bios, 'P', &bit_P)) { + if (bit_P.version == 2 && bit_P.length >= 0x5a) + fan = nv_ro16(bios, bit_P.offset + 0x58); + + if (fan) { + *ver = nv_ro08(bios, fan + 0); + switch (*ver) { + case 0x10: + *hdr = nv_ro08(bios, fan + 1); + *len = nv_ro08(bios, fan + 2); + *cnt = nv_ro08(bios, fan + 3); + return fan; + default: + break; + } + } + } + + return 0x0000; +...

...; +#include <subdev/bios/fan.h> + +u16 +nvbios_fan_table(struct nouveau_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len) +{ + struct bit_entry bit_P; + u16 fan = 0x0000; + + if (!bit_entry(bios, 'P', &bit_P)) { + if (bit_P.version == 2 && bit_P.length >= 0x5a) + fan = nv_ro16(bios, bit_P.offset + 0x58); + + if (fan) { + *ver = nv_ro08(bios, fan + 0); + switch (*ver) { + case 0x10: + *hdr = nv_ro08(bios, fan + 1); + *len = nv_ro08(bios, fan + 2); + *cnt = nv_ro08(bios, fan + 3); + return fan; + default: + break; + } + } + } + + return 0x0000; +...