I recall a long time back, when Linux was configured as standard for 100Hz ticks (aka "jiffies"), the counter was initialized close to wraparound instead of 0.

I recall a long time back, when Linux was configured as standard for 100Hz ticks (aka "jiffies"), the counter was initialized close to wraparound instead of 0. The result was you typically encountered "jiffy wraparound" after a few minutes of uptime. You learned whether your system was stable in this situation fairly quickly, rather than 248 (or 497) days later. Kernel developers typically don't have uptimes measured in days. Starting the counter close to wraparound increased the likelihood it was going to get code coverage.