This report compares latency results between Default FS and Base FS under otherwise
identical conditions. Eight same-column, same-condition pairs are available. Two pairs
are excluded as log files are not yet available (260527_1 and 260527_2).
Summary
There is no consistent trend for latencies on the default filesystem compared to the
smaller base filesystem.
Outlier — 260525_3 (Base FS; isolate CPUs; DDR mods; OoB OPTEE;
Customer test & -i200 -M): This file shows anomalously high max latency
relative to its Default FS counterpart (Pair 6) and was previously flagged in the
Category 1 analysis. Set aside for further investigation.
Hardware variability caveat: Tests were run across two boards. All observed
differences may include a board-to-board variability component. This caveat applies to all
comparisons in this report.
Individual Pair Results
SDK Test Pairs
Pair 1: 260518_1 (Default FS) vs 260522_2 (Base FS)
Column: SDK test |
Shared conditions: OoB OPTEE
Pair 2: 260519_1 (Default FS) vs 260522_4 (Base FS)
Column: SDK test |
Shared conditions: OP-TEE PRNG
Pair 3: 260526_4 (Default FS) vs 260524_1 (Base FS)
Column: SDK test |
Shared conditions: DDR mods; OoB OPTEE
Pair 4: 260526_2 (Default FS) vs 260524_3 (Base FS)
Column: SDK test |
Shared conditions: isolate CPUs; DDR mods; OoB OPTEE
Customer Test Pairs
Pair 5: 260525_2 (Default FS) vs 260525_1 (Base FS)
Column: Customer test |
Shared conditions: isolate CPUs; DDR mods; OoB OPTEE
Pair 7: 260523_2 (Default FS) vs 260526_1 (Base FS)
Column: Customer test |
Shared conditions: isolate CPUs; DDR mods; OP-TEE PRNG
Customer Test & -i200 -M Pairs
Pair 6: 260525_4 (Default FS) vs 260525_3 (Base FS)
Column: Customer test & -i200 -M |
Shared conditions: isolate CPUs; DDR mods; OoB OPTEE
Pair 8: 260523_1 (Default FS) vs 260526_3 (Base FS)
Column: Customer test & -i200 -M |
Shared conditions: isolate CPUs; DDR mods; OP-TEE PRNG
