to start with: Thanks Bill for taking the time to answer :-). I DO appreciate.
The only thing I really want us to agree on is that your approach is really a handles-only approach (when communicating between odp threads), regardless of what a thread is. Because if we don't define the behaviour, in other cases then it cannot be used. If we agree on that, then at least, I can say I understand your proposal. As long as you say "this is the minimum we guarentee" , and the "maximum" is not defined, I feel anyone will have to use the minimum. => handles I am afraid a handle only approach would have too poor performance on some system, yes, but this another story. This is even maybe wrong. To start with I just want to understand how your proposal could be used by an programmer On 9 June 2016 at 16:30, Bill Fischofer <bill.fischo...@linaro.org> wrote: > > > On Thu, Jun 9, 2016 at 9:01 AM, Christophe Milard > <christophe.mil...@linaro.org> wrote: >> >> On 9 June 2016 at 14:30, Bill Fischofer <bill.fischo...@linaro.org> wrote: >> > >> > >> > On Thu, Jun 9, 2016 at 7:13 AM, Christophe Milard >> > <christophe.mil...@linaro.org> wrote: >> >> >> >> Bill: >> >> S19: When you write:"These addresses are intended to be used within >> >> the scope of the calling thread and should not be assumed to have any >> >> validity outside of that context", do you really mean this, regardless >> >> on how ODP threads are implemented? >> > >> > >> > Yes. This is simply a statement of the minimum guarantees that ODP >> > makes. >> > Applications that wish to ensure maximum portability should follow this >> > practice. If they wish to make use of special knowledge outside of what >> > the >> > ODP API guarantees they are free to do so with the understanding that >> > they >> > are trading off portability for some other benefit known to them. >> >> But who is expected to provide that "special knowledge outside of what >> the ODP API guarantees" ? I guess that the fact that an packet data >> address returned by an ODP API call is valid within all odpthreads >> sharing the pool (i.e. if you fork after pool create, you are fine) >> belongs to that "special knowledge"... >> Where do you expect the user to find this kind of information if not >> in the ODP docs? >> >> How would a user know that a child process can make use of an odp >> pointer after fork? The user then have to know whether the pointer >> points to a shared page or not... These are things that the odp >> implementation knows... > > > The object you are sharing is the packet (represented by its handle), not a > pointer to some part of a packet mapping. That's the key distinction. > Separating object manipulation from object mapping is one of the key > distinguishing features of ODP that makes it possible to implement on a wide > array of platforms that have very different internal memory models. > > Pointers are needed only if you want to request unmediated access to some > part of a packet contents. If you're manipulating packet headroom or > tailroom, for example, you're using the packet handle, not pointers. If > you're splitting a packet you use packet handles, not pointers. If you're > adding, removing, or copying data from a packet you're using handles, not > pointers. The ODP APIs are handle-oriented, not pointer oriented, so that's > what you share between various parts of your application. > >> >> >> how would a user dare assuming things which are not documented? when >> using pthreads, who says odp_init_local() does not invalidate some of >> the ODP pointers? >> >> As a programmer, I'd go for the rule: undefined = don't use. >> >> following that rule would make me use handles only. Portable: yes. >> Acceptable performance wise... not sure... > > > No, that's not what we're saying. What we're saying is that packets are > represented by handles and programs communicate by passing these handles > around via event queues. If you want to look at the contents of a packet you > can obtain addressability to portions of a packet via well-defined ODP API > calls and those pointers have meaning for you. When someone else obtains > access to the packet handle it may do the same. > > What is the specific use case you're worried about? Petri mentioned wanting > to have multiple threads work on a single packet at the same time. Ignore > for a moment the rarity of such cases (the normal packet processing model is > one packet per thread and the reason why we have things like ordered queues > is so that you can process multiple packets in a single flow concurrently) > this is trivially handled today without further specification. Presumably > each thread wants to work on a different offset in the packet, so each such > thread does so by simply calling odp_packet_offset(pkt, myoffset, &seglen, > NULL) to get addressability to that part of the packet and goes about its > business. Note that there is no other way you could program this since you > cannot "compute" a proper packet address from another address because of > possible discontinuities due to segment boundaries. That's what the ODP API > does, so you just use it as intended. > >> >> >> >> >> > >> >> >> >> >> >> Jerrin: >> >> S18: your answer to S18 is confusing me. see my returned question in >> >> the doc. please answer it :-) >> >> >> >> All: I have updated the doc with a summary table and the comments from >> >> Jon Rosen. Please, check that What I wrote matches what you think. >> >> >> >> I have also added a list of possible behaviour (numbered A to F) for >> >> S19. Can you read these and pick your choice? If none of these >> >> alternative match your choice, please add your description. >> >> >> >> Christophe >> >> >> >> On 9 June 2016 at 02:53, Bill Fischofer <bill.fischo...@linaro.org> >> >> wrote: >> >> > >> >> > >> >> > On Wed, Jun 8, 2016 at 10:54 AM, Yi He <yi...@linaro.org> wrote: >> >> >> >> >> >> Hi, Christophe and team >> >> >> >> >> >> For the shmem part: >> >> >> >> >> >> S18: yes >> >> >> S19, S20: I agree with Bill's comments are very accurate and >> >> >> formalized. >> >> >> >> >> >> S21: to cover the case shm addresses(pointers) passed within data >> >> >> structure >> >> >> >> >> >> especially think of cases: >> >> >> 3) Context pointer getters (odp_queue_context(), >> >> >> odp_packet_user_ptr(), odp_timeout_user_ptr(), >> >> >> odp_tm_node_context(), odp_tm_queue_context(), and >> >> >> I agree with the solution, only feel it may meet difficulties >> >> >> in runtime: >> >> >> >> >> >> According to man page: >> >> >> Using shmat() with shmaddr equal to NULL is the preferred, >> >> >> portable way of attaching a shared memory segment. Be aware >> >> >> that the shared memory segment attached in this way may be >> >> >> attached at different addresses in different processes. >> >> >> Therefore, any pointers maintained within the shared memory >> >> >> must be made relative (typically to the starting address of >> >> >> the segment), rather than absolute. >> >> >> I found an alternate sweetheart only available in Windows, >> >> >> called based pointers (C++) >> >> >> https://msdn.microsoft.com/en-us/library/57a97k4e.aspx >> >> >> >> >> >> Maybe we can spent some time to look for a counterpart in >> >> >> std C world, that will be perfect. >> >> > >> >> > >> >> > For students of programming history, or those old enough to remember, >> >> > the >> >> > concept of BASED storage originated in the (now ancient) programming >> >> > language PL/I. Pointers to BASED storage were stored as offsets and >> >> > the >> >> > compiler automatically handled the relative addressing. They are very >> >> > convenient for this sort of purpose. >> >> > >> >> >> >> >> >> >> >> >> S23: agree with Bill's comments covered the cases. >> >> >> >> >> >> Thanks and best regards, Yi >> >> >> >> >> >> On 8 June 2016 at 17:04, Christophe Milard >> >> >> <christophe.mil...@linaro.org> >> >> >> wrote: >> >> >>> >> >> >>> OK. good that you agree (and please update the shared doc so it >> >> >>> becomes the central point of information). >> >> >>> There is something I like though, in your willingness to decouple >> >> >>> the >> >> >>> function and the pinning... >> >> >>> Even if I am not sure this can be enforced by ODP at all time (as >> >> >>> already stated), there is definitively a point in helping the >> >> >>> application that with to do so. So please keep an eye on that! >> >> >>> >> >> >>> Your opinion on S19 S20 and S21 would be very welcome as well... >> >> >>> This >> >> >>> is the main hurting point.... >> >> >>> >> >> >>> Christophe >> >> >>> >> >> >>> On 8 June 2016 at 09:41, Yi He <yi...@linaro.org> wrote: >> >> >>> > Hi, thanks Christophe and happy to discuss with and learn from >> >> >>> > team >> >> >>> > in >> >> >>> > yesterday's ARCH call :) >> >> >>> > >> >> >>> > The question which triggers this kind of thinking is: how to use >> >> >>> > ODP >> >> >>> > as >> >> >>> > a >> >> >>> > framework to produce re-usable building blocks to compose >> >> >>> > "Network >> >> >>> > Function >> >> >>> > Instance" in runtime, since until runtime it will prepare >> >> >>> > resources >> >> >>> > for >> >> >>> > function to settle down, thus comes the thought of seperating >> >> >>> > function >> >> >>> > implementation and launching. >> >> >>> > >> >> >>> > I agree your point it seems upper layer consideration, I'll have >> >> >>> > some >> >> >>> > time >> >> >>> > to gain deeper understanding and knowledge on how upstair >> >> >>> > playings, >> >> >>> > thus I >> >> >>> > agree to the current S11, S12, S13, S14, S15, S16, S17 approach >> >> >>> > and >> >> >>> > we >> >> >>> > can >> >> >>> > revisit if really realized upper layer programming practise/model >> >> >>> > affects >> >> >>> > ODP's design in this aspect. >> >> >>> > >> >> >>> > Best Regards, Yi >> >> >>> > >> >> >>> > On 7 June 2016 at 16:47, Christophe Milard >> >> >>> > <christophe.mil...@linaro.org> >> >> >>> > wrote: >> >> >>> >> >> >> >>> >> On 7 June 2016 at 10:22, Yi He <yi...@linaro.org> wrote: >> >> >>> >> > Hi, team >> >> >>> >> > >> >> >>> >> > I send my thoughts on the ODP thread part: >> >> >>> >> > >> >> >>> >> > S1, S2, S3, S4 >> >> >>> >> > Yi: Yes >> >> >>> >> > This set of statements defines ODP thread concept as a higher >> >> >>> >> > level abstraction of underlying concurrent execution context. >> >> >>> >> > >> >> >>> >> > S5, S6, S7, S8, S9, S10: >> >> >>> >> > Yi: Yes >> >> >>> >> > This set of statements add several constraints upon ODP >> >> >>> >> > instance concept. >> >> >>> >> > >> >> >>> >> > S11, S12, S13, S14, S15, S16, S17: >> >> >>> >> > Yi: Not very much >> >> >>> >> > >> >> >>> >> > Currently ODP application still mixes the Function >> >> >>> >> > Implementation >> >> >>> >> > and its Deployment, by this I mean in ODP application code, >> >> >>> >> > there >> >> >>> >> > is code to implement Function, there is also code to deploy >> >> >>> >> > the >> >> >>> >> > Function onto platform resources (CPU cores). >> >> >>> >> > >> >> >>> >> > I'd like to propose a programming practice/model as an attempt >> >> >>> >> > to >> >> >>> >> > decouple these two things, ODP application code only focuses >> >> >>> >> > on >> >> >>> >> > implementing Function, and leave it to a deployment script or >> >> >>> >> > launcher to take care the deployment with different resource >> >> >>> >> > spec >> >> >>> >> > (and sufficiency check also). >> >> >>> >> >> >> >>> >> Well, that goes straight against Barry's will that apps should >> >> >>> >> pin >> >> >>> >> their tasks on cpus... >> >> >>> >> Please join the public call today: if Barry is there you can >> >> >>> >> discuss >> >> >>> >> this: I am happy to hear other voices in this discussion >> >> >>> >> >> >> >>> >> > >> >> >>> >> > /* Use Case: a upper layer orchestrator, say Daylight is >> >> >>> >> > deploying >> >> >>> >> > * an ODP application (can be a VNF instance in my opinion) >> >> >>> >> > onto >> >> >>> >> > * some platform: >> >> >>> >> > */ >> >> >>> >> > >> >> >>> >> > /* The application can accept command line options */ >> >> >>> >> > --list-launchable >> >> >>> >> > list all algorithm functions that need to be arranged >> >> >>> >> > into an execution unit. >> >> >>> >> > >> >> >>> >> > --preparing-threads >> >> >>> >> > control<1,2,3>@cpu<1>, worker<1,2,3,4>@cpu<2,3,4,5> >> >> >>> >> > I'm not sure if the control/worker distinguish is needed >> >> >>> >> > anymore, DPDK has similar concept lcore. >> >> >>> >> > >> >> >>> >> > --wait-launching-signal >> >> >>> >> > main process can pause after preparing above threads but >> >> >>> >> > before launching algorithm functions, here orchestrator can >> >> >>> >> > further fine-tuning the threads by scripting (cgroups, etc). >> >> >>> >> > CPU, disk/net IO, memory quotas, interrupt bindings, etc. >> >> >>> >> > >> >> >>> >> > --launching >> >> >>> >> > main@control<1>,algorithm_one@control<2>, >> >> >>> >> > algorithm_two@worker<1,2,3,4>... >> >> >>> >> > >> >> >>> >> > In source code, the only thing ODP library and application >> >> >>> >> > need >> >> >>> >> > to >> >> >>> >> > do >> >> >>> >> > related to deployment is to declare launchable algorithm >> >> >>> >> > functions >> >> >>> >> > by >> >> >>> >> > adding them into special text section: >> >> >>> >> > >> >> >>> >> > int main(...) __attribute__((section(".launchable"))); >> >> >>> >> > int algorithm_one(void *) >> >> >>> >> > __attribute__((section(".launchable"))); >> >> >>> >> >> >> >>> >> interresting ... Does it need to be part of ODP, though? Cannot >> >> >>> >> the >> >> >>> >> ODP api provide means of pinning, and the apps that wish it can >> >> >>> >> provide the options you mentioned to do it: i.e. the application >> >> >>> >> that >> >> >>> >> wants would pin according to its command line interface and the >> >> >>> >> application that can't/does not want would pin manually... >> >> >>> >> I mean one could also imagine cases where the app needs to pin: >> >> >>> >> if >> >> >>> >> specific HW exists connected to some cpus, or if serialisation >> >> >>> >> exists >> >> >>> >> (e.g. something like: pin tasks 1,2,and 3 to cpu 1,2,3, and >> >> >>> >> then, >> >> >>> >> when >> >> >>> >> this work is done (e.g. after a barrier joining task 1,2,3) pin >> >> >>> >> taks >> >> >>> >> 4,5,6 on cpu 1,2,3 again.) There could be theoriticaly complex >> >> >>> >> such >> >> >>> >> dependency graphs where all cpus cannot be pinned from the >> >> >>> >> beginning, >> >> >>> >> and where your approach seem too restrictive. >> >> >>> >> >> >> >>> >> But I am glad to hear your voice on the arch call :-) >> >> >>> >> >> >> >>> >> Christophe. >> >> >>> >> > >> >> >>> >> > Best Regards, Yi >> >> >>> >> > >> >> >>> >> > >> >> >>> >> > On 4 June 2016 at 05:56, Bill Fischofer >> >> >>> >> > <bill.fischo...@linaro.org> >> >> >>> >> > wrote: >> >> >>> >> >> >> >> >>> >> >> I realized I forgot to respond to s23. Corrected here. >> >> >>> >> >> >> >> >>> >> >> On Fri, Jun 3, 2016 at 4:15 AM, Christophe Milard < >> >> >>> >> >> christophe.mil...@linaro.org> wrote: >> >> >>> >> >> >> >> >>> >> >> > since V3: Update following Bill's comments >> >> >>> >> >> > since V2: Update following Barry and Bill's comments >> >> >>> >> >> > since V1: Update following arch call 31 may 2016 >> >> >>> >> >> > >> >> >>> >> >> > This is a tentative to sum up the discussions around the >> >> >>> >> >> > thread/process >> >> >>> >> >> > that have been happening these last weeks. >> >> >>> >> >> > Sorry for the formalism of this mail, but it seems we need >> >> >>> >> >> > accuracy >> >> >>> >> >> > here... >> >> >>> >> >> > >> >> >>> >> >> > This summary is organized as follows: >> >> >>> >> >> > >> >> >>> >> >> > It is a set of statements, each of them expecting a >> >> >>> >> >> > separate >> >> >>> >> >> > answer >> >> >>> >> >> > from you. When no specific ODP version is specified, the >> >> >>> >> >> > statement >> >> >>> >> >> > regards the"ultimate" goal (i.e what we want eventually to >> >> >>> >> >> > achieve). >> >> >>> >> >> > Each statement is prefixed with: >> >> >>> >> >> > - a statement number for further reference (e.g. S1) >> >> >>> >> >> > - a status word (one of 'agreed' or 'open', or 'closed'). >> >> >>> >> >> > Agreed statements expect a yes/no answers: 'yes' meaning >> >> >>> >> >> > that >> >> >>> >> >> > you >> >> >>> >> >> > acknowledge that this is your understanding of the >> >> >>> >> >> > agreement >> >> >>> >> >> > and >> >> >>> >> >> > will >> >> >>> >> >> > not nack an implementation based on this statement. You can >> >> >>> >> >> > comment >> >> >>> >> >> > after a yes, but your comment will not block any >> >> >>> >> >> > implementation >> >> >>> >> >> > based >> >> >>> >> >> > on the agreed statement. A 'no' implies that the statement >> >> >>> >> >> > does >> >> >>> >> >> > not >> >> >>> >> >> > reflect your understanding of the agreement, or you refuse >> >> >>> >> >> > the >> >> >>> >> >> > proposal. >> >> >>> >> >> > Any 'no' received on an 'agreed' statement will push it >> >> >>> >> >> > back >> >> >>> >> >> > as >> >> >>> >> >> > 'open'. >> >> >>> >> >> > Open statements are fully open for further discussion. >> >> >>> >> >> > >> >> >>> >> >> > S1 -agreed: an ODP thread is an OS/platform concurrent >> >> >>> >> >> > execution >> >> >>> >> >> > environment object (as opposed to an ODP objects). No more >> >> >>> >> >> > specific >> >> >>> >> >> > definition is given by the ODP API itself. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S2 -agreed: Each ODP implementation must tell what is >> >> >>> >> >> > allowed >> >> >>> >> >> > to >> >> >>> >> >> > be >> >> >>> >> >> > used as ODP thread for that specific implementation: a >> >> >>> >> >> > linux-based >> >> >>> >> >> > implementation, for instance, will have to state whether >> >> >>> >> >> > odp >> >> >>> >> >> > threads >> >> >>> >> >> > can be linux pthread, linux processes, or both, or any >> >> >>> >> >> > other >> >> >>> >> >> > type >> >> >>> >> >> > of >> >> >>> >> >> > concurrent execution environment. ODP implementations can >> >> >>> >> >> > put >> >> >>> >> >> > any >> >> >>> >> >> > restriction they wish on what an ODP thread is allowed to >> >> >>> >> >> > be. >> >> >>> >> >> > This >> >> >>> >> >> > should be documented in the ODP implementation >> >> >>> >> >> > documentation. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S3 -agreed: in the linux generic ODP implementation a >> >> >>> >> >> > odpthread >> >> >>> >> >> > will >> >> >>> >> >> > be >> >> >>> >> >> > either: >> >> >>> >> >> > * a linux process descendant (or same as) the odp >> >> >>> >> >> > instantiation >> >> >>> >> >> > process. >> >> >>> >> >> > * a pthread 'member' of a linux process descendant >> >> >>> >> >> > (or >> >> >>> >> >> > same >> >> >>> >> >> > as) the odp instantiation process. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S4 -agreed: For monarch, the linux generic ODP >> >> >>> >> >> > implementation >> >> >>> >> >> > only >> >> >>> >> >> > supports odp thread as pthread member of the instantiation >> >> >>> >> >> > process. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S5 -agreed: whether multiple instances of ODP can be run >> >> >>> >> >> > on >> >> >>> >> >> > the >> >> >>> >> >> > same >> >> >>> >> >> > machine is left as a implementation decision. The ODP >> >> >>> >> >> > implementation >> >> >>> >> >> > document should state what is supported and any restriction >> >> >>> >> >> > is >> >> >>> >> >> > allowed. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S6 -agreed: The l-g odp implementation will support >> >> >>> >> >> > multiple >> >> >>> >> >> > odp >> >> >>> >> >> > instances whose instantiation processes are different and >> >> >>> >> >> > not >> >> >>> >> >> > ancestor/descendant of each others. Different instances of >> >> >>> >> >> > ODP >> >> >>> >> >> > will, >> >> >>> >> >> > of course, be restricted in sharing common OS ressources >> >> >>> >> >> > (The >> >> >>> >> >> > total >> >> >>> >> >> > amount of memory available for each ODP instances may >> >> >>> >> >> > decrease >> >> >>> >> >> > as >> >> >>> >> >> > the >> >> >>> >> >> > number of instances increases, the access to network >> >> >>> >> >> > interfaces >> >> >>> >> >> > will >> >> >>> >> >> > probably be granted to the first instance grabbing the >> >> >>> >> >> > interface >> >> >>> >> >> > and >> >> >>> >> >> > denied to others... some other rule may apply when sharing >> >> >>> >> >> > other >> >> >>> >> >> > common ODP ressources.) >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S7 -agreed: the l-g odp implementation will not support >> >> >>> >> >> > multiple >> >> >>> >> >> > ODP >> >> >>> >> >> > instances initiated from the same linux process (calling >> >> >>> >> >> > odp_init_global() multiple times). >> >> >>> >> >> > As an illustration, This means that a single process P is >> >> >>> >> >> > not >> >> >>> >> >> > allowed >> >> >>> >> >> > to execute the following calls (in any order) >> >> >>> >> >> > instance1 = odp_init_global() >> >> >>> >> >> > instance2 = odp_init_global() >> >> >>> >> >> > pthread_create (and, in that thread, run >> >> >>> >> >> > odp_local_init(instance1) ) >> >> >>> >> >> > pthread_create (and, in that thread, run >> >> >>> >> >> > odp_local_init(instance2) ) >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > ------------------- >> >> >>> >> >> > >> >> >>> >> >> > S8 -agreed: the l-g odp implementation will not support >> >> >>> >> >> > multiple >> >> >>> >> >> > ODP >> >> >>> >> >> > instances initiated from related linux processes >> >> >>> >> >> > (descendant/ancestor >> >> >>> >> >> > of each other), hence enabling ODP 'sub-instance'? As an >> >> >>> >> >> > illustration, >> >> >>> >> >> > this means that the following is not supported: >> >> >>> >> >> > instance1 = odp_init_global() >> >> >>> >> >> > pthread_create (and, in that thread, run >> >> >>> >> >> > odp_local_init(instance1) ) >> >> >>> >> >> > if (fork()==0) { >> >> >>> >> >> > instance2 = odp_init_global() >> >> >>> >> >> > pthread_create (and, in that thread, run >> >> >>> >> >> > odp_local_init(instance2) ) >> >> >>> >> >> > } >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > -------------------- >> >> >>> >> >> > >> >> >>> >> >> > S9 -agreed: the odp instance passed as parameter to >> >> >>> >> >> > odp_local_init() >> >> >>> >> >> > must always be one of the odp_instance returned by >> >> >>> >> >> > odp_global_init() >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S10 -agreed: For l-g, if the answer to S7 and S8 are 'yes', >> >> >>> >> >> > then >> >> >>> >> >> > due >> >> >>> >> >> > to >> >> >>> >> >> > S3, >> >> >>> >> >> > the odp_instance an odp_thread can attach to is completely >> >> >>> >> >> > defined by >> >> >>> >> >> > the ancestor of the thread, making the odp_instance >> >> >>> >> >> > parameter >> >> >>> >> >> > of >> >> >>> >> >> > odp_init_local redundant. The odp l-g implementation guide >> >> >>> >> >> > will >> >> >>> >> >> > enlighten this >> >> >>> >> >> > redundancy, but will stress that even in this case the >> >> >>> >> >> > parameter >> >> >>> >> >> > to >> >> >>> >> >> > odp_local_init() still have to be set correctly, as its >> >> >>> >> >> > usage >> >> >>> >> >> > is >> >> >>> >> >> > internal to the implementation. >> >> >>> >> >> > >> >> >>> >> >> > Barry: I think so >> >> >>> >> >> > Bill: This practice also ensures that applications behave >> >> >>> >> >> > unchanged >> >> >>> >> >> > if >> >> >>> >> >> > and when multi-instance support is added, so I don't think >> >> >>> >> >> > we >> >> >>> >> >> > need to >> >> >>> >> >> > be apologetic about this parameter requirement. >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S11 -agreed: at odp_global_init() time, the application >> >> >>> >> >> > will >> >> >>> >> >> > provide >> >> >>> >> >> > 3 >> >> >>> >> >> > sets of cpu (i.e 3 cpu masks): >> >> >>> >> >> > -the control cpu mask >> >> >>> >> >> > -the worker cpu mask >> >> >>> >> >> > -the odp service cpu mask (i.e the set of cpu odp >> >> >>> >> >> > can >> >> >>> >> >> > take >> >> >>> >> >> > for >> >> >>> >> >> > its own usage) >> >> >>> >> >> > Note: The service CPU mask will be introdused post monarch >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S12 -agreed: the odp implementation may return an error at >> >> >>> >> >> > odp_init_global() call if the number of cpu in the odp >> >> >>> >> >> > service >> >> >>> >> >> > mask >> >> >>> >> >> > (or their 'position') does not match the ODP implementation >> >> >>> >> >> > need. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes. However, an implementation may fail an >> >> >>> >> >> > odp_init_global() >> >> >>> >> >> > call >> >> >>> >> >> > for any resource insufficiency, not just cpus. >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S13 -agreed: the application is fully responsible of >> >> >>> >> >> > pinning >> >> >>> >> >> > its >> >> >>> >> >> > own >> >> >>> >> >> > odp threads to different cpus, and this is done directly >> >> >>> >> >> > through >> >> >>> >> >> > OS >> >> >>> >> >> > system calls, or via helper functions (as opposed to ODP >> >> >>> >> >> > API >> >> >>> >> >> > calls). >> >> >>> >> >> > This pinning should be made among cpus member of the worker >> >> >>> >> >> > cpu >> >> >>> >> >> > mask >> >> >>> >> >> > or the control cpu mask. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES, but I support the existence of helper functions >> >> >>> >> >> > to >> >> >>> >> >> > do >> >> >>> >> >> > this >> >> >>> >> >> > – including the >> >> >>> >> >> > important case of pinning the main thread >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes. And agree an ODP helper is useful here (which is >> >> >>> >> >> > why >> >> >>> >> >> > odp-linux >> >> >>> >> >> > provides one). >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S14 -agreed: whether more than one odp thread can be pinned >> >> >>> >> >> > to >> >> >>> >> >> > the >> >> >>> >> >> > same cpu is left as an implementation choice (and the >> >> >>> >> >> > answer >> >> >>> >> >> > to >> >> >>> >> >> > that >> >> >>> >> >> > question can be different for the service, worker and >> >> >>> >> >> > control >> >> >>> >> >> > threads). This choice should be well documented in the >> >> >>> >> >> > implementation >> >> >>> >> >> > user manual. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S15 -agreed: the odp implementation is responsible of >> >> >>> >> >> > pinning >> >> >>> >> >> > its >> >> >>> >> >> > own >> >> >>> >> >> > service threads among the cpu member of the odp service cpu >> >> >>> >> >> > mask. >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES, in principle – BUT be aware that currently the >> >> >>> >> >> > l-g >> >> >>> >> >> > ODP >> >> >>> >> >> > implementation >> >> >>> >> >> > (and perhaps many others) cannot call the helper functions >> >> >>> >> >> > (unless >> >> >>> >> >> > inlined), >> >> >>> >> >> > so this internal pinning may not be well coordinated with >> >> >>> >> >> > the >> >> >>> >> >> > helpers. >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes. And I agree with Barry on the helper recursion >> >> >>> >> >> > issue. >> >> >>> >> >> > We >> >> >>> >> >> > should >> >> >>> >> >> > fix that so there is no conflict between implementation >> >> >>> >> >> > internal >> >> >>> >> >> > pinning >> >> >>> >> >> > and application pinning attempts. >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S16 -open: why does the odp implementation need to know the >> >> >>> >> >> > control >> >> >>> >> >> > and >> >> >>> >> >> > worker mask? If S13 is true, shoudln't these two masks be >> >> >>> >> >> > part >> >> >>> >> >> > of >> >> >>> >> >> > the >> >> >>> >> >> > helper only? (meaning that S11 is wrong) >> >> >>> >> >> > >> >> >>> >> >> > Barry: Currently it probably doesn’t NEED them, but perhaps >> >> >>> >> >> > in >> >> >>> >> >> > the >> >> >>> >> >> > future, with some >> >> >>> >> >> > new API’s and capabilities, it might benefit from this >> >> >>> >> >> > information, >> >> >>> >> >> > and so I would leave them in. >> >> >>> >> >> > >> >> >>> >> >> > Bill: The implementation sees these because they are how a >> >> >>> >> >> > provisioning >> >> >>> >> >> > agent (e.g., OpenDaylight) would pass higher-level >> >> >>> >> >> > configuration >> >> >>> >> >> > information through the application to the underlying ODP >> >> >>> >> >> > implementation. >> >> >>> >> >> > The initial masks specified on odp_init_global() are used >> >> >>> >> >> > in >> >> >>> >> >> > the >> >> >>> >> >> > implementation of the odp_cpumask_default_worker(), >> >> >>> >> >> > odp_cpumask_default_control(), and >> >> >>> >> >> > odp_cpumask_all_available() >> >> >>> >> >> > APIs. >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S17 -open: should masks passed as parameter to >> >> >>> >> >> > odp_init_global() >> >> >>> >> >> > have >> >> >>> >> >> > the >> >> >>> >> >> > same "namespace" as those used internally within ODP? >> >> >>> >> >> > >> >> >>> >> >> > Barry: YES >> >> >>> >> >> > Bill: Yes. I'm not sure what it would mean for them to be >> >> >>> >> >> > in a >> >> >>> >> >> > different >> >> >>> >> >> > namespace. How would those be bridged if they weren't? >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> > --------------------------- >> >> >>> >> >> > >> >> >>> >> >> > S18 -agreed: ODP handles are valid over the whole ODP >> >> >>> >> >> > instance, >> >> >>> >> >> > i.e. >> >> >>> >> >> > any odp handle remains valid among all the odpthreads of >> >> >>> >> >> > the >> >> >>> >> >> > ODP >> >> >>> >> >> > instance regardless of the odp thread type (process, thread >> >> >>> >> >> > or >> >> >>> >> >> > whatever): an ODP thread A can pass an odp handle to >> >> >>> >> >> > onother >> >> >>> >> >> > ODP >> >> >>> >> >> > thread B (using any kind of IPC), and B can use the handle. >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes >> >> >>> >> >> > >> >> >>> >> >> > ----------------- >> >> >>> >> >> > >> >> >>> >> >> > S19 -open : any pointer retrieved by an ODP call (such as >> >> >>> >> >> > odp_*_get_addr()) follows the rules defined by the OS, with >> >> >>> >> >> > the >> >> >>> >> >> > possible exception defined in S21. For the linux generic >> >> >>> >> >> > ODP >> >> >>> >> >> > implementation, this means that >> >> >>> >> >> > pointers are fully shareable when using pthreads and that >> >> >>> >> >> > pointers >> >> >>> >> >> > pointing to shared mem areas will be shareable as long as >> >> >>> >> >> > the >> >> >>> >> >> > fork() >> >> >>> >> >> > happens after the shm_reserve(). >> >> >>> >> >> > >> >> >>> >> >> > Barry: NO. Disagree. I would prefer to see a consistent >> >> >>> >> >> > ODP >> >> >>> >> >> > answer >> >> >>> >> >> > on >> >> >>> >> >> > this topic, and in >> >> >>> >> >> > particular I don’t even believe that most OS’s “have rules >> >> >>> >> >> > defining >> >> >>> >> >> > …”, >> >> >>> >> >> > since >> >> >>> >> >> > in fact one can make programs run under Linux which can >> >> >>> >> >> > share >> >> >>> >> >> > pointers >> >> >>> >> >> > regardless >> >> >>> >> >> > the ordering of fork() calls. Most OS have lots of >> >> >>> >> >> > (continually >> >> >>> >> >> > evolving) capabilities >> >> >>> >> >> > in the category of sharing memory and so “following the >> >> >>> >> >> > rules >> >> >>> >> >> > of >> >> >>> >> >> > the >> >> >>> >> >> > OS” >> >> >>> >> >> > is not >> >> >>> >> >> > well defined. >> >> >>> >> >> > Instead, I prefer a simpler rule. Memory reserved using >> >> >>> >> >> > the >> >> >>> >> >> > special >> >> >>> >> >> > flag is guaranteed >> >> >>> >> >> > to use the same addresses across processes, and all other >> >> >>> >> >> > pointers >> >> >>> >> >> > are >> >> >>> >> >> > not guaranteed >> >> >>> >> >> > to be the same nor guaranteed to be different, so the ODP >> >> >>> >> >> > programmer >> >> >>> >> >> > should avoid >> >> >>> >> >> > any such assumptions for maximum portability. But of >> >> >>> >> >> > course >> >> >>> >> >> > programmers >> >> >>> >> >> > often >> >> >>> >> >> > only consider a subset of possible targets (e.g. how many >> >> >>> >> >> > programmers >> >> >>> >> >> > consider porting >> >> >>> >> >> > to an 8-bit CPU or a machine with a 36-bit word length), >> >> >>> >> >> > and >> >> >>> >> >> > so >> >> >>> >> >> > they >> >> >>> >> >> > may happily take advantage >> >> >>> >> >> > of certain non-guaranteed assumptions. >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> > Bill: As I noted earlier we have to distinguish between >> >> >>> >> >> > different >> >> >>> >> >> > types >> >> >>> >> >> > of >> >> >>> >> >> > memory and where these pointers come from. If the >> >> >>> >> >> > application >> >> >>> >> >> > is >> >> >>> >> >> > using >> >> >>> >> >> > malloc() or some other OS API to get memory and then using >> >> >>> >> >> > that >> >> >>> >> >> > memory's >> >> >>> >> >> > address as, for example, a queue context pointer, then it >> >> >>> >> >> > is >> >> >>> >> >> > taking >> >> >>> >> >> > responsibility for ensuring that these pointers are >> >> >>> >> >> > meaningful >> >> >>> >> >> > to >> >> >>> >> >> > whoever >> >> >>> >> >> > sees them. ODP isn't going to do anything to help there. So >> >> >>> >> >> > this >> >> >>> >> >> > question >> >> >>> >> >> > really only refers to addresses returned from ODP APIs. If >> >> >>> >> >> > we >> >> >>> >> >> > look >> >> >>> >> >> > for >> >> >>> >> >> > void >> >> >>> >> >> > * returns in the ODP API we see that the only addresses ODP >> >> >>> >> >> > returns >> >> >>> >> >> > are: >> >> >>> >> >> > >> >> >>> >> >> > 1) Those that enable addressability to buffers and packets >> >> >>> >> >> > (odp_buffer_addr(), odp_packet_data(), odp_packet_offset(), >> >> >>> >> >> > etc.) >> >> >>> >> >> > >> >> >>> >> >> > These addresses are intended to be used within the scope of >> >> >>> >> >> > the >> >> >>> >> >> > calling >> >> >>> >> >> > thread and should not be assumed to have any validity >> >> >>> >> >> > outside >> >> >>> >> >> > of >> >> >>> >> >> > that >> >> >>> >> >> > context because the buffer/packet is the durable object and >> >> >>> >> >> > any >> >> >>> >> >> > addresses >> >> >>> >> >> > are just (potentially transient) mappings of that object >> >> >>> >> >> > for >> >> >>> >> >> > use >> >> >>> >> >> > by >> >> >>> >> >> > that >> >> >>> >> >> > thread. Packet and buffer handles (not addresses) are >> >> >>> >> >> > passed >> >> >>> >> >> > between >> >> >>> >> >> > threads via queues and the receiver issues its own such >> >> >>> >> >> > calls >> >> >>> >> >> > on >> >> >>> >> >> > the >> >> >>> >> >> > received handles to get its own addressability to these >> >> >>> >> >> > objects. >> >> >>> >> >> > Whether >> >> >>> >> >> > or >> >> >>> >> >> > not these addresses are the same is purely internal to an >> >> >>> >> >> > ODP >> >> >>> >> >> > implementation and is not visible to the application. >> >> >>> >> >> > >> >> >>> >> >> > 2) Packet user areas (odp_packet_user_area()). This API >> >> >>> >> >> > returns >> >> >>> >> >> > the >> >> >>> >> >> > address of a preallocated user area associated with the >> >> >>> >> >> > packet >> >> >>> >> >> > (size >> >> >>> >> >> > defined by the pool that the packet was drawn from at >> >> >>> >> >> > odp_pool_create() >> >> >>> >> >> > time by the max_uarea_size entry in the odp_pool_param_t). >> >> >>> >> >> > Since >> >> >>> >> >> > this >> >> >>> >> >> > is >> >> >>> >> >> > metadata associated with the packet this API may be called >> >> >>> >> >> > by >> >> >>> >> >> > any >> >> >>> >> >> > thread >> >> >>> >> >> > that obtains the odp_packet_t for the packet that contains >> >> >>> >> >> > that >> >> >>> >> >> > user >> >> >>> >> >> > area. >> >> >>> >> >> > However, like the packet itself, the scope of this returned >> >> >>> >> >> > address >> >> >>> >> >> > is >> >> >>> >> >> > the >> >> >>> >> >> > calling thread. So the address returned by >> >> >>> >> >> > odp_packet_user_area() >> >> >>> >> >> > should >> >> >>> >> >> > not be cached or passed to any other thread. Each thread >> >> >>> >> >> > that >> >> >>> >> >> > needs >> >> >>> >> >> > addressability to this area makes its own call and whether >> >> >>> >> >> > these >> >> >>> >> >> > returned >> >> >>> >> >> > addresses are the same or different is again internal to >> >> >>> >> >> > the >> >> >>> >> >> > implementation >> >> >>> >> >> > and not visible to the application. Note that just as two >> >> >>> >> >> > threads >> >> >>> >> >> > should >> >> >>> >> >> > not ownership of an odp_packet_t at the same time, two >> >> >>> >> >> > threads >> >> >>> >> >> > should >> >> >>> >> >> > not >> >> >>> >> >> > be trying to access the user area associated with a packet >> >> >>> >> >> > either. >> >> >>> >> >> > >> >> >>> >> >> > 3) Context pointer getters (odp_queue_context(), >> >> >>> >> >> > odp_packet_user_ptr(), >> >> >>> >> >> > odp_timeout_user_ptr(), odp_tm_node_context(), >> >> >>> >> >> > odp_tm_queue_context(), >> >> >>> >> >> > and >> >> >>> >> >> > the user context pointer carried in the odp_crypto_params_t >> >> >>> >> >> > struct) >> >> >>> >> >> > >> >> >>> >> >> > These are set by the application using corresponding setter >> >> >>> >> >> > APIs >> >> >>> >> >> > or >> >> >>> >> >> > provided as values in structs, so the application either >> >> >>> >> >> > obtains >> >> >>> >> >> > these >> >> >>> >> >> > pointers on its own, in which case it is responsible for >> >> >>> >> >> > ensuring >> >> >>> >> >> > that >> >> >>> >> >> > they >> >> >>> >> >> > are meaningful to whoever retrieves them, or from an >> >> >>> >> >> > odp_shm_t. >> >> >>> >> >> > So >> >> >>> >> >> > these >> >> >>> >> >> > are not a special case in themselves. >> >> >>> >> >> > >> >> >>> >> >> > 4) ODP shared memory (odp_shm_addr(), odp_shm_info()). >> >> >>> >> >> > These >> >> >>> >> >> > APIs >> >> >>> >> >> > return >> >> >>> >> >> > addresses to odp_shm_t objects that are specifically >> >> >>> >> >> > created >> >> >>> >> >> > to >> >> >>> >> >> > support >> >> >>> >> >> > sharing. The rule here is simple: the scope of any returned >> >> >>> >> >> > shm >> >> >>> >> >> > address >> >> >>> >> >> > is >> >> >>> >> >> > determined by the sharing flag specified at >> >> >>> >> >> > odp_shm_reserve() >> >> >>> >> >> > time. >> >> >>> >> >> > ODP >> >> >>> >> >> > currently defines two such flags: ODP_SHM_SW_ONLY and >> >> >>> >> >> > ODP_SHM_PROC. >> >> >>> >> >> > We >> >> >>> >> >> > simply need to define precisely the intended sharing scope >> >> >>> >> >> > of >> >> >>> >> >> > these >> >> >>> >> >> > two >> >> >>> >> >> > (or >> >> >>> >> >> > any new flags we define) to answer this question. Note >> >> >>> >> >> > that >> >> >>> >> >> > context >> >> >>> >> >> > pointers drawn from odp_shm_t objects would then have >> >> >>> >> >> > whatever >> >> >>> >> >> > sharing >> >> >>> >> >> > attributes that the shm object has, thus completely >> >> >>> >> >> > defining >> >> >>> >> >> > case >> >> >>> >> >> > (3). >> >> >>> >> >> > >> >> >>> >> >> > --------------------- >> >> >>> >> >> > >> >> >>> >> >> > S20 -open: by default, shmem addresses (returned by >> >> >>> >> >> > odp_shm_addr()) >> >> >>> >> >> > follow the OS rules, as defined by S19. >> >> >>> >> >> > >> >> >>> >> >> > Ola: The question is which OS rules apply (an OS can have >> >> >>> >> >> > different >> >> >>> >> >> > rules >> >> >>> >> >> > for >> >> >>> >> >> > different OS objects, e.g. memory regions allocated using >> >> >>> >> >> > malloc >> >> >>> >> >> > and >> >> >>> >> >> > mmap >> >> >>> >> >> > will behave differently). I think the answer depends on ODP >> >> >>> >> >> > shmem >> >> >>> >> >> > objects >> >> >>> >> >> > are implemented. Only the ODP implementation knows how ODP >> >> >>> >> >> > shmem >> >> >>> >> >> > objects >> >> >>> >> >> > are created (e.g. use some OS system call, manipulate the >> >> >>> >> >> > page >> >> >>> >> >> > tables >> >> >>> >> >> > directly). So essentially the sharability of pointers is >> >> >>> >> >> > ODP >> >> >>> >> >> > implementation >> >> >>> >> >> > specific (although ODP implementations on the same OS can >> >> >>> >> >> > be >> >> >>> >> >> > expected >> >> >>> >> >> > to >> >> >>> >> >> > behave the same). Conclusion: we actually don't specify >> >> >>> >> >> > anything >> >> >>> >> >> > at >> >> >>> >> >> > all >> >> >>> >> >> > here, it is completely up to the ODP implementation. >> >> >>> >> >> > What is required/expected by ODP applications? If we don't >> >> >>> >> >> > make >> >> >>> >> >> > applications happy, ODP is unlikely to succeed. >> >> >>> >> >> > I think many applications are happy with a single-process >> >> >>> >> >> > thread >> >> >>> >> >> > model >> >> >>> >> >> > where all memory is shared and pointers can be shared >> >> >>> >> >> > freely. >> >> >>> >> >> > I hear of some applications that require multi-process >> >> >>> >> >> > thread >> >> >>> >> >> > model, >> >> >>> >> >> > I >> >> >>> >> >> > expect that those applications also want to be able to >> >> >>> >> >> > share >> >> >>> >> >> > memory >> >> >>> >> >> > and >> >> >>> >> >> > pointers freely between them, at least memory that was >> >> >>> >> >> > specifically >> >> >>> >> >> > allocated to be shared (so called shared memory regions, >> >> >>> >> >> > what's >> >> >>> >> >> > otherwise >> >> >>> >> >> > the purpose of such memory regions?). >> >> >>> >> >> > >> >> >>> >> >> > Barry: Disagree with the same comments as in S19. >> >> >>> >> >> > >> >> >>> >> >> > Bill: I believe my discourse on S19 completely resolves >> >> >>> >> >> > this >> >> >>> >> >> > question. >> >> >>> >> >> > This >> >> >>> >> >> > is controlled by the share flag specified at >> >> >>> >> >> > odp_shm_reserve() >> >> >>> >> >> > time. >> >> >>> >> >> > We >> >> >>> >> >> > just need to specify the sharing scope implied by each of >> >> >>> >> >> > these >> >> >>> >> >> > and >> >> >>> >> >> > then >> >> >>> >> >> > it >> >> >>> >> >> > is up to each implementation to see that such scope is >> >> >>> >> >> > realized. >> >> >>> >> >> > >> >> >>> >> >> > --------------------- >> >> >>> >> >> > >> >> >>> >> >> > S21 -open: shm will support and extra flag at shm_reserve() >> >> >>> >> >> > call >> >> >>> >> >> > time: >> >> >>> >> >> > SHM_XXX. The usage of this flag will allocate shared memory >> >> >>> >> >> > guaranteed >> >> >>> >> >> > to be located at the same virtual address on all odpthreads >> >> >>> >> >> > of >> >> >>> >> >> > the >> >> >>> >> >> > odp_instance. Pointers to this shared memory type are >> >> >>> >> >> > therefore >> >> >>> >> >> > fully >> >> >>> >> >> > sharable, even on odpthreads running on different VA space >> >> >>> >> >> > (e.g. >> >> >>> >> >> > processes). The amount of memory which can be allocated >> >> >>> >> >> > using >> >> >>> >> >> > this >> >> >>> >> >> > flag can be >> >> >>> >> >> > limited to any value by the ODP implementation, down to >> >> >>> >> >> > zero >> >> >>> >> >> > bytes, >> >> >>> >> >> > meaning that some odp implementation may not support this >> >> >>> >> >> > option >> >> >>> >> >> > at >> >> >>> >> >> > all. The shm_reserve() will return an error in this >> >> >>> >> >> > case.The >> >> >>> >> >> > usage of >> >> >>> >> >> > this flag by the application is therefore not recommended. >> >> >>> >> >> > The >> >> >>> >> >> > ODP >> >> >>> >> >> > implementation may require a hint about the size of this >> >> >>> >> >> > area >> >> >>> >> >> > at >> >> >>> >> >> > odp_init_global() call time. >> >> >>> >> >> > >> >> >>> >> >> > Barry: Mostly agree, except for the comment about the >> >> >>> >> >> > special >> >> >>> >> >> > flag >> >> >>> >> >> > not >> >> >>> >> >> > being recommended. >> >> >>> >> >> > >> >> >>> >> >> > Ola: Agree. Some/many applications will want to share >> >> >>> >> >> > memory >> >> >>> >> >> > between >> >> >>> >> >> > threads/processes and must be able to do so. Some ODP >> >> >>> >> >> > platforms >> >> >>> >> >> > may >> >> >>> >> >> > have >> >> >>> >> >> > limitations to the amount of memory (if any) that can be >> >> >>> >> >> > shared >> >> >>> >> >> > and >> >> >>> >> >> > may >> >> >>> >> >> > thus fail to run certain applications. Such is life. I >> >> >>> >> >> > don't >> >> >>> >> >> > see >> >> >>> >> >> > a >> >> >>> >> >> > problem >> >> >>> >> >> > with that. Possibly we should remove the phrase "not >> >> >>> >> >> > recommended" >> >> >>> >> >> > and >> >> >>> >> >> > just >> >> >>> >> >> > state that portability may be limited. >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> > Bill: Yes. As noted in S19 and S20 the intent of the share >> >> >>> >> >> > flag >> >> >>> >> >> > is to >> >> >>> >> >> > specify desired addressability scope for the returned >> >> >>> >> >> > odp_shm_t. >> >> >>> >> >> > It's >> >> >>> >> >> > perfectly reasonable to define multiple such scopes that >> >> >>> >> >> > may >> >> >>> >> >> > have >> >> >>> >> >> > different >> >> >>> >> >> > intended uses (and implementation costs). >> >> >>> >> >> > >> >> >>> >> >> > ------------------ >> >> >>> >> >> > >> >> >>> >> >> > S22 -open: please put here your name suggestions for this >> >> >>> >> >> > SHM_XXX >> >> >>> >> >> > flag >> >> >>> >> >> > :-). >> >> >>> >> >> > >> >> >>> >> >> > Ola: >> >> >>> >> >> > SHM_I_REALLY_WANT_TO_SHARE_THIS_MEMORY >> >> >>> >> >> > >> >> >>> >> >> > Bill: I previously suggested ODP_SHM_INSTANCE that >> >> >>> >> >> > specifies >> >> >>> >> >> > that >> >> >>> >> >> > the >> >> >>> >> >> > sharing scope of this odp_shm_t is the entire ODP instance. >> >> >>> >> >> > >> >> >>> >> >> > ------------------ >> >> >>> >> >> > >> >> >>> >> >> > S23 -open: The rules above define relatively well the >> >> >>> >> >> > behaviour >> >> >>> >> >> > of >> >> >>> >> >> > pointer retrieved by the call to odp_shm_get_addr(). But >> >> >>> >> >> > many >> >> >>> >> >> > points >> >> >>> >> >> > needs tobe defined regarding other ODP objects pointers: >> >> >>> >> >> > What >> >> >>> >> >> > is >> >> >>> >> >> > the >> >> >>> >> >> > validity of a pointer to a packet, for instance? If process >> >> >>> >> >> > A >> >> >>> >> >> > creates >> >> >>> >> >> > a packet pool P, then forks B and C, and B allocate a >> >> >>> >> >> > packet >> >> >>> >> >> > from >> >> >>> >> >> > P >> >> >>> >> >> > and retrieves a pointer to a packet allocated from this >> >> >>> >> >> > P... >> >> >>> >> >> > Is >> >> >>> >> >> > this >> >> >>> >> >> > pointer valid in A and C? In the current l-g >> >> >>> >> >> > implementation, >> >> >>> >> >> > it >> >> >>> >> >> > will... Is this behaviour >> >> >>> >> >> > something we wish to enforce on any odp implementation? >> >> >>> >> >> > What >> >> >>> >> >> > about >> >> >>> >> >> > other objects: buffers, atomics ... Some clear rule has to >> >> >>> >> >> > be >> >> >>> >> >> > defined >> >> >>> >> >> > here... How things behave and if this behaviour is a part >> >> >>> >> >> > of >> >> >>> >> >> > the >> >> >>> >> >> > ODP >> >> >>> >> >> > API or just specific to different implementations... >> >> >>> >> >> > >> >> >>> >> >> > Ola: Perhaps we need the option to specify the >> >> >>> >> >> > I_REALLY_WANT_TO_SHARE_THIS_MEMORY flag when creating all >> >> >>> >> >> > types >> >> >>> >> >> > of >> >> >>> >> >> > ODP >> >> >>> >> >> > pools? >> >> >>> >> >> > An ODP implementation can always fail to create such a pool >> >> >>> >> >> > if >> >> >>> >> >> > the >> >> >>> >> >> > sharability requirement can not be satisfied. >> >> >>> >> >> > Allocation of locations used for atomic operations is the >> >> >>> >> >> > responsibility >> >> >>> >> >> > of the application which can (and must) choose a suitable >> >> >>> >> >> > type >> >> >>> >> >> > of >> >> >>> >> >> > memory. >> >> >>> >> >> > It is better that sharability is an explicit requirement >> >> >>> >> >> > from >> >> >>> >> >> > the >> >> >>> >> >> > application. It should be specified as a flag parameter to >> >> >>> >> >> > the >> >> >>> >> >> > different >> >> >>> >> >> > calls that create/allocate regions of memory (shmem, >> >> >>> >> >> > different >> >> >>> >> >> > types >> >> >>> >> >> > of >> >> >>> >> >> > pools). >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> > Barry: >> >> >>> >> >> > Again refer to S19 answer. Specifically it is about what >> >> >>> >> >> > is >> >> >>> >> >> > GUARANTEED regarding >> >> >>> >> >> > pointer validity, not whether the pointers in certain cases >> >> >>> >> >> > will >> >> >>> >> >> > happen >> >> >>> >> >> > to be >> >> >>> >> >> > the same. So for your example, the pointer is not >> >> >>> >> >> > guaranteed >> >> >>> >> >> > to >> >> >>> >> >> > be >> >> >>> >> >> > valid in A and C, >> >> >>> >> >> > but the programmer might well believe that for all the ODP >> >> >>> >> >> > platforms >> >> >>> >> >> > and implementations >> >> >>> >> >> > they expect to run on, this is very likely to be the case, >> >> >>> >> >> > in >> >> >>> >> >> > which >> >> >>> >> >> > case we can’t stop them >> >> >>> >> >> > from constraining their program’s portability – no more >> >> >>> >> >> > than >> >> >>> >> >> > requiring >> >> >>> >> >> > them to be able to >> >> >>> >> >> > port to a ternary (3-valued “bit”) architecture. >> >> >>> >> >> > >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> Bill: See s19 response, which answers all these questions. >> >> >>> >> >> The >> >> >>> >> >> fork >> >> >>> >> >> case >> >> >>> >> >> you mention is moot because addresses returned by ODP packet >> >> >>> >> >> operations >> >> >>> >> >> are >> >> >>> >> >> only valid in the thread that makes those calls. The handle >> >> >>> >> >> must >> >> >>> >> >> be >> >> >>> >> >> valid >> >> >>> >> >> throughout the instance so that may affect how the >> >> >>> >> >> implementation >> >> >>> >> >> chooses >> >> >>> >> >> to implement packet pool creation, but such questions are >> >> >>> >> >> internal >> >> >>> >> >> to >> >> >>> >> >> each >> >> >>> >> >> implementation and not part of the API. >> >> >>> >> >> >> >> >>> >> >> Atomics are application declared entities and so memory >> >> >>> >> >> sharing >> >> >>> >> >> is >> >> >>> >> >> again >> >> >>> >> >> covered by my response to s19. If the atomic is in storage >> >> >>> >> >> the >> >> >>> >> >> application >> >> >>> >> >> allocated from the OS, then its sharability is the >> >> >>> >> >> responsibility >> >> >>> >> >> of >> >> >>> >> >> the >> >> >>> >> >> application. If the atomic is part of a struct that resides >> >> >>> >> >> in >> >> >>> >> >> an >> >> >>> >> >> odp_shm_t, then its sharability is governed by the share flag >> >> >>> >> >> of >> >> >>> >> >> the >> >> >>> >> >> odp_shm it is taken from. >> >> >>> >> >> >> >> >>> >> >> No additional parameters are required for pools because what >> >> >>> >> >> is >> >> >>> >> >> shared >> >> >>> >> >> from >> >> >>> >> >> them is handles, not addresses. As noted, when these handles >> >> >>> >> >> are >> >> >>> >> >> converted >> >> >>> >> >> to addresses those addresses are only valid for the calling >> >> >>> >> >> thread. >> >> >>> >> >> All >> >> >>> >> >> other threads that have access to the handle make their own >> >> >>> >> >> calls >> >> >>> >> >> and >> >> >>> >> >> whether or not those two addresses have any relationship to >> >> >>> >> >> each >> >> >>> >> >> other >> >> >>> >> >> is >> >> >>> >> >> not visible to the application. >> >> >>> >> >> >> >> >>> >> >> >> >> >>> >> >> > >> >> >>> >> >> >> >> >>> >> >> > --------------------- >> >> >>> >> >> > >> >> >>> >> >> > Thanks for your feedback! >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> > >> >> >>> >> >> _______________________________________________ >> >> >>> >> >> lng-odp mailing list >> >> >>> >> >> lng-odp@lists.linaro.org >> >> >>> >> >> https://lists.linaro.org/mailman/listinfo/lng-odp >> >> >>> >> > >> >> >>> >> > >> >> >>> > >> >> >>> > >> >> >> >> >> >> >> >> > >> > >> > > > _______________________________________________ lng-odp mailing list lng-odp@lists.linaro.org https://lists.linaro.org/mailman/listinfo/lng-odp