Alexander Motin 6210ac95a1 amd64: Stop using REP MOVSB for backward memmove()s. ... Enhanced REP MOVSB feature of CPUs starting from Ivy Bridge makes REP MOVSB the fastest way to copy memory in most of cases. However Intel Optimization Reference Manual says: "setting the DF to force REP MOVSB to copy bytes from high towards low addresses will expe- rience significant performance degradation". Measurements on Intel Cascade Lake and Alder Lake, same as on AMD Zen3 show that it can drop throughput to as low as 2.5-3.5GB/s, comparing to ~10-30GB/s of REP MOVSQ or hand-rolled loop, used for non-ERMS CPUs. This patch keeps ERMS use for forward ordered memory copies, but removes it for backward overlapped moves where it does not work. Reviewed by: mjg MFC after: 2 weeks		2022-06-16 13:46:34 -04:00
..
acpica	amd64: Remove unused devclass arguments to DRIVER_MODULE.	2022-05-06 15:46:59 -07:00
amd64	amd64: Stop using REP MOVSB for backward memmove()s.	2022-06-16 13:46:34 -04:00
conf	ktls: Remove the KERN_TLS option from the i386 and amd64 LINT-NOIP kernel configurations.	2022-06-11 21:31:28 +02:00
ia32	x86: Add a NT_X86_SEGBASES register set.	2022-03-24 11:36:19 -07:00
include	kasan: Create a shadow for the bootstack prior to hammer_time()	2022-06-15 11:39:10 -04:00
linux	linux(4): Properly build argument list for the signal handler	2022-05-30 19:53:12 +03:00
linux32	linux(4): Properly restore the thread signal mask after signal delivery on i386	2022-05-30 20:03:49 +03:00
pci	pci_cfgreg.c: Use io port config access for early boot time.	2019-04-09 18:07:17 +00:00
sgx	sgx: Remove unused variable.	2022-04-07 17:01:28 -07:00
vmm	vmm: move bumping VMEXIT_USERSPACE stat to the right place	2022-06-09 08:57:25 -04:00
Makefile	Bring the tags and links entries for amd64 up to date.	2021-09-27 20:04:51 -07:00