* [DPDK/core Bug 1385] rt_bitops.h fails to give implied atomicity guarantees
@ 2024-02-26 9:56 bugzilla
0 siblings, 0 replies; only message in thread
From: bugzilla @ 2024-02-26 9:56 UTC (permalink / raw)
To: dev
[-- Attachment #1: Type: text/plain, Size: 1506 bytes --]
https://bugs.dpdk.org/show_bug.cgi?id=1385
Bug ID: 1385
Summary: rt_bitops.h fails to give implied atomicity guarantees
Product: DPDK
Version: 23.11
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Priority: Normal
Component: core
Assignee: dev@dpdk.org
Reporter: mattias.ronnblom@ericsson.com
Target Milestone: ---
The documentation (and the naming) for the rte_bit_relaxed_*() functions in
rte_bitops.h makes clear that all such functions have a relaxed memory order.
The use of the term "relaxed", which most C programmers likely are familiar
with from the C11 memory model specification, itself implies that the
operations are supposed to be atomic. Why otherwise mention the memory
operations are relaxed? Relaxed is the default for non-atomic loads and stores.
In addition, why otherwise declare the address as volatile?
An even stronger indication are the test-and-set family of "relaxed"
rte_bitops.h functions. "Test-and-set" in the low-level C programming context
always implies an atomic operation. A non-atomic test-and-set does not make
sense.
In summary, a perfectly valid interpretation of the API contract is that the
functions are atomic.
However, their implementation is not, which may not be obvious to the causal
API user.
--
You are receiving this mail because:
You are the assignee for the bug.
[-- Attachment #2: Type: text/html, Size: 3418 bytes --]
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2024-02-26 9:56 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-02-26 9:56 [DPDK/core Bug 1385] rt_bitops.h fails to give implied atomicity guarantees bugzilla
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).