Hi Kennon,
Some perf reports and analysis imply that backward moves (with overlap?) are no
faster than straight rep movsb on some CPUs, so it may be better to just
simplify to that, unless you want to stash the final element(s) to be moved out
of the way in register(s), and use multiple registers in unrolled wide moves for
the aligned portion?
On 2026-02-27 11:20, KENNON J CONRAD via Cygwin wrote:
I updated the BIOS to the newest version and ran the code again. The
memmove got SIG_TRAP again. This time the code was freeing up uint16_t array
element 1525 and eliminating array element 1676 by moving elements 1525 - 1675
to elements 1526 - 1676.
RSI contains the address of element 1601, indicating the rep movsq did only 9
iterations before the SIG_TRAP happened. The ZF shows a value of 0. So it
seems like this is perhaps an OS problem with ZF handling of interrupts. At
this point, it is easiest for me to avoid memmove for backward moves with
source/destination overlap because of this problem. I don't know how many
computers are impacted by this problem but I do know it's causing crashes on my
test computer and that's enough reason to avoid it.
On 02/26/2026 7:06 PM PST Eliot Moss wrote:
On 2/26/2026 9:04 PM, KENNON J CONRAD wrote:
Perhaps something like this? Maybe I need to try a BIOS update?
Yes, something *like* these flaws. The thing is to find out if your specific
processor has a flaw that needs an update, etc.
Intel Haswell (and related architectures) processors may experience stability
issues, including machine check errors (MCEs), due to a microcode bug related
to REP MOVS (specifically REP MOVSB or REP MOVSQ) handling. These issues often
cause system crashes or lockups, leading to microcode, BIOS/UEFI updates to
resolve them.
Debian Wiki
+4
Issue: A high-rate of interrupts or specific memory operations can cause REP
MOVS instructions to trigger Machine Check Errors (MCE) or internal errors
(IERR) on older processors.
Affected Processors: The bug primarily impacts older Intel processors,
including Haswell and Broadwell architectures.
Fix/Mitigation: The primary solution is to apply the latest motherboard
BIOS/UEFI update, which contains the corrected microcode update (often labelled
20180108 or later).
Background: This is linked to Intel’s efforts to fix security vulnerabilities
(like Spectre), where early patches for Haswell/Broadwell caused unexpected
instability, requiring further refinements.
Debian Wiki
+4
--
Take care. Thanks, Brian Inglis Calgary, Alberta, Canada
La perfection est atteinte Perfection is achieved
non pas lorsqu'il n'y a plus rien à ajouter not when there is no more to add
mais lorsqu'il n'y a plus rien à retrancher but when there is no more to cut
-- Antoine de Saint-Exupéry
--
Problem reports: https://cygwin.com/problems.html
FAQ: https://cygwin.com/faq/
Documentation: https://cygwin.com/docs.html
Unsubscribe info: https://cygwin.com/ml/#unsubscribe-simple
