DPDK patches and discussions
 help / color / mirror / Atom feed
From: "Mattias Rönnblom" <hofors@lysator.liu.se>
To: "Tyler Retzlaff" <roretzla@linux.microsoft.com>,
	"Mattias Rönnblom" <mattias.ronnblom@ericsson.com>
Cc: dev@dpdk.org, Heng Wang <heng.wang@ericsson.com>,
	Stephen Hemminger <stephen@networkplumber.org>
Subject: Re: [RFC v2 0/6] Improve EAL bit operations API
Date: Fri, 26 Apr 2024 13:17:40 +0200	[thread overview]
Message-ID: <923ae68e-52a4-4f0e-abb4-4940f1a796e0@lysator.liu.se> (raw)
In-Reply-To: <20240425180513.GA23608@linuxonhyperv3.guj3yctzbm1etfxqx2vob5hsef.xx.internal.cloudapp.net>

On 2024-04-25 20:05, Tyler Retzlaff wrote:
> On Thu, Apr 25, 2024 at 10:58:47AM +0200, Mattias Rönnblom wrote:
>> This patch set represent an attempt to improve and extend the RTE
>> bitops API, in particular for functions that operate on individual
>> bits.
>>
>> All new functionality is exposed to the user as generic selection
>> macros, delegating the actual work to private (__-marked) static
>> inline functions. Public functions (e.g., rte_bit_set32()) would just
>> be bloating the API. Such generic selection macros will here be
>> referred to as "functions", although technically they are not.
> 
> 
>>
>> The legacy <rte_bitops.h> rte_bit_relaxed_*() family of functions is
>> replaced with three families:
>>
>> rte_bit_[test|set|clear|assign]() which provides no memory ordering or
>> atomicity guarantees and no read-once or write-once semantics (e.g.,
>> no use of volatile), but does provide the best performance. The
>> performance degradation resulting from the use of volatile (e.g.,
>> forcing loads and stores to actually occur and in the number
>> specified) and atomic (e.g., LOCK-prefixed instructions on x86) may be
>> significant.
>>
>> rte_bit_once_*() which guarantees program-level load and stores
>> actually occurring (i.e., prevents certain optimizations). The primary
>> use of these functions are in the context of memory mapped
>> I/O. Feedback on the details (semantics, naming) here would be greatly
>> appreciated, since the author is not much of a driver developer.
>>
>> rte_bit_atomic_*() which provides atomic bit-level operations,
>> including the possibility to specifying memory ordering constraints
>> (or the lack thereof).
>>
>> The atomic functions take non-_Atomic pointers, to be flexible, just
>> like the GCC builtins and default <rte_stdatomic.h>. The issue with
>> _Atomic APIs is that it may well be the case that the user wants to
>> perform both non-atomic and atomic operations on the same word.
>>
>> Having _Atomic-marked addresses would complicate supporting atomic
>> bit-level operations in the bitset API (proposed in a different RFC
>> patchset), and potentially other APIs depending on RTE bitops for
>> atomic bit-level ops). Either one needs two bitset variants, one
>> _Atomic bitset and one non-atomic one, or the bitset code needs to
>> cast the non-_Atomic pointer to an _Atomic one. Having a separate
>> _Atomic bitset would be bloat and also prevent the user from both, in
>> some situations, doing atomic operations against a bit set, while in
>> other situations (e.g., at times when MT safety is not a concern)
>> operating on the same objects in a non-atomic manner.
> 
> understood. i think the only downside is that if you do have an
> _Atomic-specified type you'll have to cast the qualification away
> to use the function like macro.
> 

This is tricky, and I can't say I've really converged on an opinion, but 
it seems to me at this point you shouldn't mark anything _Atomic.

> as a suggestion the _Generic legs could include both _Atomic-specified
> and non-_Atomic-specified types where an intermediate inline function
> could strip the qualification to use your core inline implementations.
> 
> _Generic((v), int *: __foo32, RTE_ATOMIC(int) *: __foo32_unqual)(v))
> 
> static inline void
> __foo32(int *a) { ... }
> 
> static inline void
> __foo32_unqual(RTE_ATOMIC(int) *a) { __foo32((int *)(uintptr_t)(a)); }
> 
> there is some similar prior art in newer ISO C23 with typeof_unqual.
> 
> https://en.cppreference.com/w/c/language/typeof
> 

This is an interesting solution, but I'm not sure it's a problem that 
needs to be solved.

>>
>> Unlike rte_bit_relaxed_*(), individual bits are represented by bool,
>> not uint32_t or uint64_t. The author found the use of such large types
>> confusing, and also failed to see any performance benefits.
>>
>> A set of functions rte_bit_*_assign() are added, to assign a
>> particular boolean value to a particular bit.
>>
>> All new functions have properly documented semantics.
>>
>> All new functions (or more correctly, generic selection macros)
>> operate on both 32 and 64-bit words, with type checking.
>>
>> _Generic allow the user code to be a little more impact. Have a
>> type-generic atomic test/set/clear/assign bit API also seems
>> consistent with the "core" (word-size) atomics API, which is generic
>> (both GCC builtins and <rte_stdatomic.h> are).
> 
> ack, can you confirm _Generic is usable from a C++ TU? i may be making a
> mistake locally but using g++ version 11.4.0 -std=c++20 it wasn't
> accepting it.
> 
> i think using _Generic is ideal, it eliminates mistakes when handling
> the different integer sizes so if it turns out C++ doesn't want to
> cooperate the function like macro can conditionally expand to a C++
> template this will need to be done for MSVC since i can confirm
> _Generic does not work with MSVC C++.
> 

That's unfortunate.

No, I didn't try it with C++. I just assumed _Generic was C++ as well.

The naive solution would be to include two overloaded functions per 
function-like macro.

#ifdef __cplusplus

#undef rte_bit_set

static inline void
rte_bit_set(uint32_t *addr, unsigned int nr)
{
     __rte_bit_set32(addr, nr);
}

static inline void
rte_bit_set(uint64_t *addr, unsigned int nr)
{
     __rte_bit_set64(addr, nr);
}
#endif

Did you have something more clever/less verbose in mind? The best would 
if one could have a completely generic C++-compatible replacement of 
_Generic, but it's not obvious how that would work.

What's the minimum C++ version required by DPDK? C++11?

>>
>> The _Generic versions avoids having explicit unsigned long versions of
>> all functions. If you have an unsigned long, it's safe to use the
>> generic version (e.g., rte_set_bit()) and _Generic will pick the right
>> function, provided long is either 32 or 64 bit on your platform (which
>> it is on all DPDK-supported ABIs).
>>
>> The generic rte_bit_set() is a macro, and not a function, but
>> nevertheless has been given a lower-case name. That's how C11 does it
>> (for atomics, and other _Generic), and <rte_stdatomic.h>. Its address
>> can't be taken, but it does not evaluate its parameters more than
>> once.
>>
>> Things that are left out of this patch set, that may be included
>> in future versions:
>>
>>   * Have all functions returning a bit number have the same return type
>>     (i.e., unsigned int).
>>   * Harmonize naming of some GCC builtin wrappers (i.e., rte_fls_u32()).
>>   * Add __builtin_ffsll()/ffs() wrapper and potentially other wrappers
>>     for useful/used bit-level GCC builtins.
>>   * Eliminate the MSVC #ifdef-induced documentation duplication.
>>   * _Generic versions of things like rte_popcount32(). (?)
> 
> it would be nice to see them all converted, at the time i added them we
> still hadn't adopted C11 so was limited. but certainly not asking for it
> as a part of this series.
> 
>>
>> Mattias Rönnblom (6):
>>    eal: extend bit manipulation functionality
>>    eal: add unit tests for bit operations
>>    eal: add exactly-once bit access functions
>>    eal: add unit tests for exactly-once bit access functions
>>    eal: add atomic bit operations
>>    eal: add unit tests for atomic bit access functions
>>
>>   app/test/test_bitops.c       | 319 +++++++++++++++++-
>>   lib/eal/include/rte_bitops.h | 624 ++++++++++++++++++++++++++++++++++-
>>   2 files changed, 925 insertions(+), 18 deletions(-)
>>
>> -- 
> 
> Series-acked-by: Tyler Retzlaff <roretzla@linux.microsoft.com>
> 
>> 2.34.1

  reply	other threads:[~2024-04-26 11:17 UTC|newest]

Thread overview: 90+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-02 13:53 [RFC 0/7] " Mattias Rönnblom
2024-03-02 13:53 ` [RFC 1/7] eal: extend bit manipulation functions Mattias Rönnblom
2024-03-02 17:05   ` Stephen Hemminger
2024-03-03  6:26     ` Mattias Rönnblom
2024-03-04 16:34       ` Tyler Retzlaff
2024-03-05 18:01         ` Mattias Rönnblom
2024-03-05 18:06           ` Tyler Retzlaff
2024-04-25  8:58   ` [RFC v2 0/6] Improve EAL bit operations API Mattias Rönnblom
2024-04-25  8:58     ` [RFC v2 1/6] eal: extend bit manipulation functionality Mattias Rönnblom
2024-04-29  9:51       ` [RFC v3 0/6] Improve EAL bit operations API Mattias Rönnblom
2024-04-29  9:51         ` [RFC v3 1/6] eal: extend bit manipulation functionality Mattias Rönnblom
2024-04-29 11:12           ` Morten Brørup
2024-04-30  9:55           ` [RFC v4 0/6] Improve EAL bit operations API Mattias Rönnblom
2024-04-30  9:55             ` [RFC v4 1/6] eal: extend bit manipulation functionality Mattias Rönnblom
2024-04-30 12:08               ` [RFC v5 0/6] Improve EAL bit operations API Mattias Rönnblom
2024-04-30 12:08                 ` [RFC v5 1/6] eal: extend bit manipulation functionality Mattias Rönnblom
2024-05-02  5:57                   ` [RFC v6 0/6] Improve EAL bit operations API Mattias Rönnblom
2024-05-02  5:57                     ` [RFC v6 1/6] eal: extend bit manipulation functionality Mattias Rönnblom
2024-05-05  8:37                       ` [RFC v7 0/6] Improve EAL bit operations API Mattias Rönnblom
2024-05-05  8:37                         ` [RFC v7 1/6] eal: extend bit manipulation functionality Mattias Rönnblom
2024-05-05  8:37                         ` [RFC v7 2/6] eal: add unit tests for bit operations Mattias Rönnblom
2024-05-05  8:37                         ` [RFC v7 3/6] eal: add exactly-once bit access functions Mattias Rönnblom
2024-05-07 19:17                           ` Morten Brørup
2024-05-08  6:47                             ` Mattias Rönnblom
2024-05-08  7:33                               ` Morten Brørup
2024-05-08  8:00                                 ` Mattias Rönnblom
2024-05-08  8:11                                   ` Morten Brørup
2024-05-08  9:27                                     ` Mattias Rönnblom
2024-05-08 10:08                                       ` Morten Brørup
2024-05-08 15:15                                 ` Stephen Hemminger
2024-05-08 16:16                                   ` Morten Brørup
2024-05-05  8:37                         ` [RFC v7 4/6] eal: add unit tests for " Mattias Rönnblom
2024-05-05  8:37                         ` [RFC v7 5/6] eal: add atomic bit operations Mattias Rönnblom
2024-05-05  8:37                         ` [RFC v7 6/6] eal: add unit tests for atomic bit access functions Mattias Rönnblom
2024-05-02  5:57                     ` [RFC v6 2/6] eal: add unit tests for bit operations Mattias Rönnblom
2024-05-02  5:57                     ` [RFC v6 3/6] eal: add exactly-once bit access functions Mattias Rönnblom
2024-05-02  5:57                     ` [RFC v6 4/6] eal: add unit tests for " Mattias Rönnblom
2024-05-02  5:57                     ` [RFC v6 5/6] eal: add atomic bit operations Mattias Rönnblom
2024-05-03  6:41                       ` Mattias Rönnblom
2024-05-03 23:30                         ` Tyler Retzlaff
2024-05-04 15:36                           ` Mattias Rönnblom
2024-05-02  5:57                     ` [RFC v6 6/6] eal: add unit tests for atomic bit access functions Mattias Rönnblom
2024-04-30 12:08                 ` [RFC v5 2/6] eal: add unit tests for bit operations Mattias Rönnblom
2024-04-30 12:08                 ` [RFC v5 3/6] eal: add exactly-once bit access functions Mattias Rönnblom
2024-04-30 12:08                 ` [RFC v5 4/6] eal: add unit tests for " Mattias Rönnblom
2024-04-30 12:08                 ` [RFC v5 5/6] eal: add atomic bit operations Mattias Rönnblom
2024-04-30 12:08                 ` [RFC v5 6/6] eal: add unit tests for atomic bit access functions Mattias Rönnblom
2024-04-30  9:55             ` [RFC v4 2/6] eal: add unit tests for bit operations Mattias Rönnblom
2024-04-30  9:55             ` [RFC v4 3/6] eal: add exactly-once bit access functions Mattias Rönnblom
2024-04-30  9:55             ` [RFC v4 4/6] eal: add unit tests for " Mattias Rönnblom
2024-04-30 10:37               ` Morten Brørup
2024-04-30 11:58                 ` Mattias Rönnblom
2024-04-30  9:55             ` [RFC v4 5/6] eal: add atomic bit operations Mattias Rönnblom
2024-04-30  9:55             ` [RFC v4 6/6] eal: add unit tests for atomic bit access functions Mattias Rönnblom
2024-04-29  9:51         ` [RFC v3 2/6] eal: add unit tests for bit operations Mattias Rönnblom
2024-04-29  9:51         ` [RFC v3 3/6] eal: add exactly-once bit access functions Mattias Rönnblom
2024-04-29  9:51         ` [RFC v3 4/6] eal: add unit tests for " Mattias Rönnblom
2024-04-29  9:51         ` [RFC v3 5/6] eal: add atomic bit operations Mattias Rönnblom
2024-04-29  9:51         ` [RFC v3 6/6] eal: add unit tests for atomic bit access functions Mattias Rönnblom
2024-04-25  8:58     ` [RFC v2 2/6] eal: add unit tests for bit operations Mattias Rönnblom
2024-04-25  8:58     ` [RFC v2 3/6] eal: add exactly-once bit access functions Mattias Rönnblom
2024-04-25  8:58     ` [RFC v2 4/6] eal: add unit tests for " Mattias Rönnblom
2024-04-25  8:58     ` [RFC v2 5/6] eal: add atomic bit operations Mattias Rönnblom
2024-04-25 10:25       ` Morten Brørup
2024-04-25 14:36         ` Mattias Rönnblom
2024-04-25 16:18           ` Morten Brørup
2024-04-26  9:39             ` Mattias Rönnblom
2024-04-26 12:00               ` Morten Brørup
2024-04-28 15:37                 ` Mattias Rönnblom
2024-04-29  7:24                   ` Morten Brørup
2024-04-30 16:52               ` Tyler Retzlaff
2024-04-25  8:58     ` [RFC v2 6/6] eal: add unit tests for atomic bit access functions Mattias Rönnblom
2024-04-25 18:05     ` [RFC v2 0/6] Improve EAL bit operations API Tyler Retzlaff
2024-04-26 11:17       ` Mattias Rönnblom [this message]
2024-04-26 21:35     ` Patrick Robb
2024-03-02 13:53 ` [RFC 2/7] eal: add generic bit manipulation macros Mattias Rönnblom
2024-03-04  8:16   ` Heng Wang
2024-03-04 15:41     ` Mattias Rönnblom
2024-03-04 16:42   ` Tyler Retzlaff
2024-03-05 18:08     ` Mattias Rönnblom
2024-03-05 18:22       ` Tyler Retzlaff
2024-03-05 20:02         ` Mattias Rönnblom
2024-03-05 20:53           ` Tyler Retzlaff
2024-03-02 13:53 ` [RFC 3/7] eal: add bit manipulation functions which read or write once Mattias Rönnblom
2024-03-02 13:53 ` [RFC 4/7] eal: add generic once-type bit operations macros Mattias Rönnblom
2024-03-02 13:53 ` [RFC 5/7] eal: add atomic bit operations Mattias Rönnblom
2024-03-02 13:53 ` [RFC 6/7] eal: add generic " Mattias Rönnblom
2024-03-02 13:53 ` [RFC 7/7] eal: deprecate relaxed family of " Mattias Rönnblom
2024-03-02 17:07   ` Stephen Hemminger
2024-03-03  6:30     ` Mattias Rönnblom

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=923ae68e-52a4-4f0e-abb4-4940f1a796e0@lysator.liu.se \
    --to=hofors@lysator.liu.se \
    --cc=dev@dpdk.org \
    --cc=heng.wang@ericsson.com \
    --cc=mattias.ronnblom@ericsson.com \
    --cc=roretzla@linux.microsoft.com \
    --cc=stephen@networkplumber.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).