Some thoughts that appeared to me when discussing the currently implemented solution for the "sigalarm fix" in pumptools: The current way this is implemented by inlining it into the main thread might be causing performance issues on weaker hardware, e.g. stuttering. Furthermore, the polling rate is likely to be lower than on non-patched code which can result in negative impact on gameplay, e.g. "laggy" weird input.
This needs to be investigated. A potential solution direction could be to look into multi-threading the function that was executed by the sigalarm handler on stock software and hardware.
Some thoughts that appeared to me when discussing the currently implemented solution for the "sigalarm fix" in pumptools: The current way this is implemented by inlining it into the main thread might be causing performance issues on weaker hardware, e.g. stuttering. Furthermore, the polling rate is likely to be lower than on non-patched code which can result in negative impact on gameplay, e.g. "laggy" weird input.
This needs to be investigated. A potential solution direction could be to look into multi-threading the function that was executed by the sigalarm handler on stock software and hardware.