nv-tegra-mirror/linux-nvgpu
1
0
Fork 0
mirror of git://nv-tegra.nvidia.com/linux-nvgpu.git synced 2025-12-24 10:34:43 +03:00
Code Issues Packages Projects Releases Wiki Activity

gpu: nvgpu: use nvgpu_udelay and optimize do while loop

Browse Source 
- For interrupt context, use nvgpu_udelay()
  instead of nvgpu_usleep_range()
- Use while loop instead of do while loop for polling
  as this will help minimize overall delay

Bug 200350539

Change-Id: Icc05911082877df7f9e807da8f16e4d54e620cd8
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1573303
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This commit is contained in:
Seema Khowala
2017-10-04 15:01:32 -07:00
committed by mobile promotions
parent 82207a3d3c
commit f606195542
1 changed files with 12 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

22
drivers/gpu/nvgpu/gk20a/priv_ring_gk20a.c
View File

@@ -38,6 +38,8 @@ void gk20a_enable_priv_ring(struct gk20a *g)
if (nvgpu_is_enabled(g, NVGPU_IS_FMODEL))
return;
nvgpu_log(g, gpu_dbg_info, "enabling priv ring");
if (g->ops.clock_gating.slcg_priring_load_gating_prod)
g->ops.clock_gating.slcg_priring_load_gating_prod(g,
g->slcg_enabled);
@@ -47,9 +49,7 @@ void gk20a_enable_priv_ring(struct gk20a *g)
gk20a_writel(g, pri_ringstation_sys_decode_config_r(),
0x2);
gk20a_readl(g, pri_ringstation_sys_decode_config_r());
}
void gk20a_priv_ring_isr(struct gk20a *g)
@@ -86,18 +86,20 @@ void gk20a_priv_ring_isr(struct gk20a *g)
gk20a_readl(g, pri_ringstation_gpc_gpc0_priv_error_code_r() + gpc * gpc_stride));
}
}
/* clear interrupt */
cmd = gk20a_readl(g, pri_ringmaster_command_r());
cmd = set_field(cmd, pri_ringmaster_command_cmd_m(),
pri_ringmaster_command_cmd_ack_interrupt_f());
gk20a_writel(g, pri_ringmaster_command_r(), cmd);
do {
/* poll for clear interrupt done */
cmd = pri_ringmaster_command_cmd_v(
gk20a_readl(g, pri_ringmaster_command_r()));
while (cmd != pri_ringmaster_command_cmd_no_cmd_v() && retry) {
nvgpu_udelay(20);
retry--;
cmd = pri_ringmaster_command_cmd_v(
gk20a_readl(g, pri_ringmaster_command_r()));
nvgpu_usleep_range(20, 40);
} while (cmd != pri_ringmaster_command_cmd_no_cmd_v() && --retry);
if (retry <= 0)
nvgpu_warn(g, "priv ringmaster cmd ack too many retries");
}
if (retry == 0 && cmd != pri_ringmaster_command_cmd_no_cmd_v())
nvgpu_warn(g, "priv ringmaster intr ack too many retries");
}