3nigm4 Posted March 11, 2015 Share Posted March 11, 2015 Hello, I have a "little" issue with my system:- HP Elite 7000 (motherboard MS 7613) (H57 chipset)- Intel i5 750- 8GB RAM (4x Kingston 2GB DDR3 1333MHz)- Yosemite 10.10.2 (iMac 11.3 SMC)9 boot on 10, I get a fully working system except the kernel_task, which is costantly using 67% of CPU: analyzing a bit, I can see a costant wake up from idle by Vector 0x49, even in safe mode.I initially used Clover patching, that worked flawlessly with any other hack I built, but trying to solve the issue I moved to a custom DSDT (the one attached), with almost any patch needed except audio and graphics (I still patch them by clover).1 boot on 10, the system runs flowlessly but I am unable to see any substantial difference in both kernel and boot logs.I think the issue may be related to USB ports 'cause I have a bad reading from an external UPS and (I don't think if it's normal) I have only 2x "USB High-speed Bus" in system profiler and no USB Buses (my old system was reporting UHCx controllers).I hope to solve this with your help... (I found that vector 0x49 on real macs may be an issue related to some faulty sensor)Than you in advancel0v3r DSDT.aml.zip Link to comment Share on other sites More sharing options...
artur_pt Posted March 11, 2015 Share Posted March 11, 2015 hello i change ur usb .. if u want can test this one DSDT.aml.zip good hack Link to comment Share on other sites More sharing options...
3nigm4 Posted March 11, 2015 Author Share Posted March 11, 2015 Thank you, I'll try it tomorrow Tried, also changing ID's, removing HUB configuration inside both EHCx: I always get the same result. Maybe is right to have only EHCI controllers in system profiler, basically all ports are working. The only issue is high interrupts on (by?) vector 0x49 of kernel_task, that once out of 10 boots disapper. I also tried a new install on another HDD with no success.... I'm thinking that is really something related to a harware issue. Link to comment Share on other sites More sharing options...
3nigm4 Posted March 23, 2015 Author Share Posted March 23, 2015 Solved! After digging around I've found this script http://www.brendangregg.com/DTrace/hotkernel from this page http://www.brendangregg.com/dtrace.html with which I've found that the kernel is doing a lot of read/write/query ACPI with AppleACPIPlatform.kext: ... AppleACPIPlatform`AcpiPsParseLoop 170 0.1% AppleACPIPlatform`AcpiPsGetNextNamepath 196 0.1% AppleACPIPlatform`AcpiUtAcquireMutex 210 0.1% kernel`0xffffff801d56fef0 210 0.1% AppleACPIPlatform`AcpiPsGetNextSimpleArg 219 0.1% AppleACPIPlatform`AcpiUtStrlen 239 0.2% AppleACPIPlatform`AcpiNsSearchOneScope 298 0.2% AppleACPIPlatform`AcpiHwValidateIoRequest 316 0.2% kernel`0xffffff801d53a870 379 0.3% kernel`ml_set_interrupts_enabled 382 0.3% AppleACPIPlatform`AcpiPsPeekOpcode 392 0.3% kernel`OSAddAtomic64 408 0.3% AppleACPIPlatform`AcpiNsLookup 436 0.3% AppleACPIPlatform`AcpiTbCheckDsdtHeader 474 0.3% kernel`0xffffff801d570350 477 0.3% kernel`lck_mtx_lock_spin 486 0.3% kernel`zfree 574 0.4% kernel`lck_mtx_unlock 586 0.4% kernel`lck_mtx_lock 596 0.4% kernel`0xffffff801d56f470 606 0.4% kernel`0xffffff801d562140 1041 0.7% AppleACPIPlatform`AcpiOsReadPort 1299 0.9% AppleACPIPlatform`AcpiUtMemset 1452 1.0% AppleACPIPlatform`AcpiUtMatchPredefinedMethod 1567 1.0% AppleACPIPlatform`AcpiOsWritePort 3174 2.1% kernel`0xffffff801d548260 16699 11.2% kernel`machine_idle 109949 73.7% I've than used following boot flags to debug ACPI: -v debug=0x12a acpi_layer=0x08 acpi_level=0x02 Upon reboot, the log started to repeatedly report SIOH related messages preventing the boot to proceed: ACPI: (SIOH) I've than rebooted without flags and edited DSDT removing SIOH and other PS/2 related blocks (PS2K, PS2M, SIOK, SIOW, SIOS) and rebooted: the system is now perfect! The MS 7613 motherboard comes with Fintek F71858 chip for monitoring the system: this chip also includes a PS/2 controller, but no PS/2 ports exists on the back, so having them in DSDT is really useless. I hope to help others in avoiding loosing so much time in solving this same problem. Cheers DSDT.aml.zip Link to comment Share on other sites More sharing options...
mrmacdi2 Posted October 25, 2023 Share Posted October 25, 2023 On 3/23/2015 at 10:49 AM, 3nigm4 said: After digging around I've found this script http://www.brendangregg.com/DTrace/hotkernel from this page http://www.brendangregg.com/dtrace.html with which I've found that the kernel is doing a lot of read/write/query ACPI with AppleACPIPlatform.kext: I have the same issue in Ventura 13.6, Sleep then Wake fixes the issue. Trying to debug using dtrace and your script, but I get this: Quote dtrace: system integrity protection is on, some features will not be available but SIP is disabled. hotkernel: Spoiler ... 0xffffff800c5a51c5 437 0.1% 0xffffff800d423b0b 474 0.1% 0xffffff800d423a0a 482 0.1% 0xffffff800c4c9379 497 0.1% 0xffffff800c4d7b8b 528 0.1% 0xffffff800d4239f1 651 0.1% 0xffffff800c53681e 680 0.1% 0xffffff800c4d8592 1001 0.1% 0xffffff800c412daa 1163 0.2% 0xffffff800c4cacc7 1267 0.2% 0xffffff800c4d7562 1484 0.2% 0xffffff800c5b6273 2010 0.3% 0xffffff800c1010f6 2817 0.4% 0xffffff800c412d8a 3156 0.4% 0xffffff800c49c0f1 4869 0.6% 0xffffff800c4a34f2 12283 1.6% 0xffffff800c5ae981 18528 2.4% 0xffffff800c5b82da 643960 83.3% Link to comment Share on other sites More sharing options...
Slice Posted October 25, 2023 Share Posted October 25, 2023 Dtrace is controlled by SIP bit 0x20 #define CSR_ALLOW_UNRESTRICTED_DTRACE (1 << 5) Remember, SIP is not only enabled or disabled, there are variants. Link to comment Share on other sites More sharing options...
Recommended Posts