PR/58775 CVS commit: [netbsd-10] src/sys/dev/acpi

["Martin Husemann" <martin%netbsd.org@localhost>]

The following reply was made to PR kern/58775; it has been noted by GNATS.

From: "Martin Husemann" <martin%netbsd.org@localhost>
To: gnats-bugs%gnats.NetBSD.org@localhost
Cc: 
Subject: PR/58775 CVS commit: [netbsd-10] src/sys/dev/acpi
Date: Fri, 1 Nov 2024 14:45:36 +0000

 Module Name:	src
 Committed By:	martin
 Date:		Fri Nov  1 14:45:36 UTC 2024
 
 Modified Files:
 	src/sys/dev/acpi [netbsd-10]: apei.c apei_cper.h apei_hest.c
 
 Log Message:
 Pull up following revision(s) (requested by riastradh in ticket #994):
 
 	sys/dev/acpi/apei_cper.h: revision 1.3
 	sys/dev/acpi/apei_cper.h: revision 1.4
 	sys/dev/acpi/apei_cper.h: revision 1.5
 	sys/dev/acpi/apei_hest.c: revision 1.4
 	sys/dev/acpi/apei_hest.c: revision 1.5
 	sys/dev/acpi/apei_hest.c: revision 1.6
 	sys/dev/acpi/apei.c: revision 1.4
 	sys/dev/acpi/apei.c: revision 1.5
 	sys/dev/acpi/apei.c: revision 1.6
 	sys/dev/acpi/apei.c: revision 1.7
 	sys/dev/acpi/apei.c: revision 1.8
 	sys/dev/acpi/apei.c: revision 1.9
 
 apei(4): Paranoia: Clamp multiplication to SIZE_MAX too.
 
 This makes it clear that the result is guaranteed not to overflow
 size_t.  Previously it was only implied because on all NetBSD ports,
 SIZE_MAX > INT32_MAX, but let's make it clearer instead of relying on
 tacitly on that assumption.
 
 No functional change intended.
 
 Noticed while preparing for:
 PR kern/58775: apei(4) spamming console
 
 apei(4): Use __SIMPLELOCK_UNLOCKED to initialize __cpu_simple_lock_t.
 Zero-initialization is wrong on some architectures.  While it happens
 to work on x86 (and this is under ifdef x86), let's just do the right
 thing to make it clearer.
 
 Noticed while preparing for:
 PR kern/58775: apei(4) spamming console
 
 apei(4): Paranoia: Don't schedule callout if poll interval is zero.
 Prompted by making sure there isn't a runaway callout in:
 PR kern/58775: apei(4) spamming console
 
 apei(4): Fix format of CPER UUID strings to match UEFI spec.
 PR kern/58775: apei(4) spamming console
 apei(4): Rate-limit error reports to console.
 
 For each severity level (corrected, uncorrectable/recoverable,
 uncorrectable/fatal, other), print at most one report per minute,
 and, if any reports were suppressed, the number of reports that had
 been suppressed since the last one, up to a maximum of 2^32 - 1.
 PR kern/58775: apei(4) spamming console
 
 apei(4): Fix comments citing structure member names.
 No functional change intended.
 In preparation for further changes for:
 PR kern/58775: apei(4) spamming console
 
 apei(4): Add section numbering in comments on error types.
 This will make it easier to keep the sections sorted by the spec's
 section numbers when we add more.
 Preparation for:
 PR kern/58775: apei(4) spamming console
 
 apei(4): Print PCIe errors.
 PR kern/58775: apei(4) spamming console
 
 apei(4): Print PCIe cap and AER extcap registers in 32-bit chunks.
 PR kern/58775: apei(4) spamming console
 
 apei(4): Print non-fatal uncorrectable PCIe AER errors as intended.
 Noted by rillig@.
 PR kern/58775: apei(4) spamming console
 
 
 To generate a diff of this commit:
 cvs rdiff -u -r1.3.4.2 -r1.3.4.3 src/sys/dev/acpi/apei.c \
     src/sys/dev/acpi/apei_hest.c
 cvs rdiff -u -r1.2.4.2 -r1.2.4.3 src/sys/dev/acpi/apei_cper.h
 
 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.