Dynamic power management in a processor design
View Patent ↗A pipeline stage's stall detection logic detects a stall condition, and sends a signal to idle detection logic to gate off the pipeline's register clocks. The stall detection logic also monitors a downstream pipeline stage's stall condition, and instructs the idle detection logic to gate off the pipeline stage's registers when the downstream pipeline stage is in a stall condition as well. In addition, when the pipeline stage's stall detection logic detects a stall condition, either from the downstream pipeline stage or from its own pipeline units, the pipeline stage's stall detection logic informs an upstream pipeline stage to gate off its clocks and thus, conserve more power.
1. A computer-implemented method comprising:
detecting a stall condition within a pipeline stage from a plurality of pipeline stages, wherein the pipeline stage includes a first register, a second register, a first fill detector, and a second fill detector, the first fill detector controlling the first register and the second fill detector controlling the second register;
activating a stall signal to the first fill detector in response to detecting the stall condition;
in response to receiving the activated stall signal, waiting until the first register includes a first instruction and gating off a first clock to the first register using the first fill detector once the first register includes the first instruction;
after gating off the first clock, sending a first register loaded signal from the first fill detector to the second fill detector; and
in response to receiving the first register loaded signal, waiting until the second register includes a second instruction and gating off a second clock to the second register using the second fill detector once the second register includes the second instruction.
2. The method of claim 1 further comprising:
detecting a downstream stall condition in a downstream pipeline stage that is included in the plurality of pipeline stages; and activating the stall signal in response to detecting the downstream stall condition.
3. The method of claim 2 further comprising: determining that the downstream stall condition has concluded; and deactivating the stall signal in response to the determination.
4. The method of claim 1 further comprising:
propagating the stall signal to an upstream pipeline stage that is included in the plurality of pipeline stages; and
wherein the propagation results in the upstream pipeline stage gating off an upstream pipeline stage clock to one or more registers included in the upstream pipeline stage.
5. The method of claim 1 wherein the stall condition is selected from the group consisting of an ERAT miss, a cache miss, a data dependency, and a non-pipeline instruction.
6. The method of claim 1 wherein a downstream pipeline stage included in the plurality of pipeline stages is not gating off a downstream pipeline stage clock to its corresponding registers while the pipeline stage is gating off the first clock and the second clock to the first register and the second register, respectively.
7. The method of claim 1 further comprising:
wherein the method is performed using a broadband processor architecture, the broadband processor architecture including a plurality of heterogeneous processors, a common memory, and a common bus; and
wherein the plurality of heterogeneous processors use different instruction sets and share the common memory and the common bus.