Questions about the load processing of OpenSparc T1 Dcache

Hi,
I have some questions about OpenSparc T1 Dcache load processing.
During load processing, subsequent loads to the same address need to search the store buffer for a valid store to that address. If there is a CAM hit, data is sourced from the store buffer, not from the D-cache, and no load request will be sent to the L2.
What if there is no CAM hit. Would the load request be sent to L2? Or would Dcache be checked for the requested data?
If the load request would be sent to L2, what next? Would the Dcache be updated?
Thanks

Store buffer is checked for Read after Write (RAW) condition on loads. If there is full RAW - i.e. full data exists in the store buffer - then the data is bypassed and no D cache access happens.
If RAW is partial (e.g. word store followed by a double word load) then load is treated as a miss. Store is allowed to complete in L2 cache and then load instruction is completed.
For the miss in STB, D cache is accessed. If hit, data is fetched from D$. If miss, data is fetched from L2$ and allocated in D$.

Similar Messages

  • A question about the Installation process in general

    Hey guys, a general question to the installation process.
    The following scenario:
    I install a programme on my macbook pro and don't tick all possible installation options (like additional content).
    Afterwards I change my mind and want to install the optional content.
    So I run the installation programme again and now the question:
    If I install the additional content AND the mainprogramme again is the main programme going to be overwritten or am I going to have the mainprogramme twice on my system from now on
    Thanks in advance for you help, folks!

    If you do a Unicode conversion (no matter in a UC & CU scenario or as a single project) the target system must be empty - means, you will create a new database and load it with the content you exported.
    You can't install the export into your already existing database with the same name.
    Markus

  • HCM Processes & Forms - Some questions about the PD Processes

    Hi Experts,
    At my Organisation we use HCM Processes and Forms and Organisation Management and MSS to manage Position Based Security (PBS).
    The MSS Managers have requested a form be developed which they can use to extend an existing Position to Employee temporary assignment (IT1001 S-P with A081 relationship) which is due to expire within the next 30 days. We recently upgraded to ECC EHP6 and I thought I would now have a go at developing this Form.
    While researching in SCN and SAP Press books on the ‘HCM Processes and Forms’ topic, I have discovered a lot of information that has assisted me so far in progressing this solution.
    There are a few givens that I have confirmed so far, which are as follows;
    Since we are using Positions (Organisation Management), the only option available here is to use the PD Process, which implies using the SAP_PD Back_End Service. Therefore this become the baseline to build upon.
    While researching in SCN and SAP Press books on this subject I could not find relevant information about what is possible using the SAP_PD Service to achieve my solution design for the above requirements and thus my post to the HCM P&F Community for assistance.
    My questions are as follows;
    In the solution design, I was considering to use the “Process without Start Object” or “Mass Start” to skip Step 1 - “Selection Object” and use a Generic Service to get all the positions for the Manager (along the lines of the SAP Standard Generic Service - find Positions in Manager’s area of Responsibility) and only retain the Positions where the End Date for the IT1001 A081 relationship in between the Effective Date and Effective Date plus 30 days. This should provide me with all the positions under the Manager that have the A081 relationship with the assigned employee that are about to expire within the next 30 days.
    I want the result presented on the Form in a Table format so that the Manager has to enter the Extension Start Date and End Date for the new assignment period.
    Is this possible using Standard Design Time functionality?
    2.       Or Instead of Question 1, could I use OADP to achieve the same result using a custom search class to return a list of multiple objects (Positions)which is used to populate the form?
    3.       I get the impression that the ‘Mass Start’ process is mainly for multiple employees. Can it also be used for multiple PD objects?
    4.       I get the impression that if I used the ‘Mass Start’ option I could only use this if I was CREATING a new object or assignment? (I take it that you cannot use the Mass Start option to CHANGE multiple existing objects. Only because I have not seen any SAP Standard examples of this yet!)   Please confirm if this is possible?
    5.       Has anyone used the “Query for Mass Start” option on a Process? How does that work? (Could I write a ‘SAP Query’ to get the positions I am interested in under the Manager?) Is there a sample SAP Standard process that uses a query?
    Your assistance would be greatly appreciated.
    Thanks in Advance.
    Regards,
    Steve

    no limit
    no limit
    limitations of what?
    you can use forms on mac, pc, and other systems

  • Question about the Signal Processing Toolkit

    I am using Measurement Studio with Visual C++ 6. Is it possible to use the
    Signal Processing Toolkit in this environment or does LabWindowsCVI have to
    be used? Would the toolkit supply a lib/dll that could be linked into a
    VC++ project and called directly from the program?
    Thanks

    You can use the signal processing toolkit libraries in VC++, but the libraries do depend on CVI libraries. The easiest way to create a Visual C++ project that links to the CVI libraries would be to either a.) create the project in CVI then convert it using the Measurement Studio CVI conversion wizard, or b.) create a Visual C++ project that uses the CVI libraries with teh Measurement Studio application wizard.
    For more information on the conversion wizard, see the following topic in the Measurement Studio references:
    (Measurement Studio for Visual C++ 6)
    Visual C++ Help\Tutorials\Getting Started\Measurement Studio for Visual C++ 6.0\Converting LabWindows/CVI Projects to Measurement Studio for Visual C++ Projects
    (Measurement Studio for Visual C++ .NET)
    Visual C++ Help\Tutorials\Getting Started\Measurement Studio 6.0 for Visual C++ .NET\Working with LabWindows/CVI Projects in Visual C++
    - Elton

  • A simple question about the Integration process and the PI system

    Hi all
    Out PI system gets more and more slowly .
    in the Monitor I can se that there are many ufinished IP about 100.
    in the Outbound colon there is a Watch icon with the message
    QMessage being sent                     
    and in the graphic of the IP thers is a Send Message Asynchronously  with text in process.
    my question is can those unfineshed IP be guilty in slowing down the system by taking resources, they have been there for quit a long time .
    Thanks

    Hi again
    I have checked sm12 , no locks
    theres is no mapping involved in this step , the IP is trying to send a file to  server.
    I just want to know if  100 unfinished IP slow down the system .
    It does not matter if the file has been sent or not
    is there any way to check those IP if they are taking resources.
    Thanks.

  • Question about the import process in UC&CU upgrade

    Hi,
    We have are doing an UC&CU upgrade ( 4.7 => ERP6 ehp4)
    We have export the database (unicode converssion) and we are in the step where it's the import.
    I would to kno what i have to do in this SAPinst step :
    Database Schema Database objects have been found in the database schema (dv1 in DV1) used for this installation.You must delete these objects, or use another database or SAP SID as entered in the previous dialogs.
    and there is check box : Delete existing database objects.
    I would like to be sure about this option.
    I supose i have to valid this option if someone could confirm this point.

    If you do a Unicode conversion (no matter in a UC & CU scenario or as a single project) the target system must be empty - means, you will create a new database and load it with the content you exported.
    You can't install the export into your already existing database with the same name.
    Markus

  • Basic Questions on the Syncing Process

    I have a couple of questions about the syncing process.
    First - does the syncing take place through MobileMe and the so called “cloud” or does it take place through the physical connection of the Touch to the iMac via the USB cable? When I connect the Touch to the iMac, I have
    sync Address Book contacts
    and
    sync iCal calendars checked in iTunes
    and there’s a little message in the same iTunes window under Bookmarks which tells me that the bookmarks are being synced over the air via MobileMe. So is the following assumption correct on my part that
    Address Book
    iCal
    are being synced via the physical USB connection while the Bookmarks are being synced over the WiFi connection?
    And finally, if I have contacts - let’s call them 1, 2, 3, 4, 5 on the Touch and only 1, 2, 3 on the iMac
    what happens with the syncing process? Do the contacts on the Touch get reduced to 1, 2, 3 or do those on the iMac expand to 1, 2, 3, 4, 5 ? The question I guess is what syncs to what? Do changes on the iMac transfer over to the Touch during syncing or does it work the other way too?
    Thanks
    John
    Thanks
    John

    1) If you have it to sync via Mobileme, then it sync via Mobileme. If you do not, then it does not.
    2) It syncs both ways.
    http://manuals.info.apple.com/enUS/iPod_touch_3.0_UserGuide.pdf

  • Questions about the content of download meeting recording .zip file

    I tried posting this on the resurrected Connect forum, but my Adobe ID wasn't recognized there....
    Concerning the files that are included in the .zip file of the meeting recording that can be downloaded:
    1) Is there any documentation describing the files and their contents (i.e. what each file represents, what each XML element and attribute in those files represent)
    2) Are there any files that capture mouse movement on a shared desktop?
    Thank you!

    Hi Sean,
    Regarding your first post:
    Thanks Jorma! I don't have access to an FMS build at the moment but I'm quite certain it's there. As for contacting Jaydeep, I am 90% sure he authorized us to broadcast his email on here if folks had questions about the tool, but, in the case that I'm wrong and he didn't - I'm going to double-check first.
    Regarding your most recent post..
    "To be clear, the most critical goal I'm trying to accomplish is to create an automated process that will download the recording meeting at its highest quality in a consistent and reliable manner".
    I personally believe this is possible; unfortunately, I haven't seen it done yet. If your recording contains:
    - audio
    - a camera feed
    - screensharing
    Then I think you might be able to get this going. If it contains shared content, like a shared PPT, this gets trickier.
    "To do this, of course, I have to reproduce some of the functionality that Connect provides, starting and combining video and audio streams according to the instructions in the control files."
    Exactly right. If your recording didn't contain shared content, then all you've got on your hands are a bunch of audio/video files that you could edit together as you wanted with your favourite video editing tool. If it contains shared content, here's (at a high level) what's happening.
    For shared PPTs or FTContent files:
    First (for version 9 recordings only), Connect reads the information on the Shared Content's location and SCO within mainstream and indexstream and validates it before loading it. I don't recall this happening to the same extent with version 8 or earlier, but maybe it was. Now, if the content is validated (ie. Connect can find it) the share pod will display as black, if it doesn't, you get an empty pod with an message like "No content is being shared" or something like that.
    Connect then looks at the actual FTContent file, and loads the content that is to be shared using the file path and sco ID listed in here. It's important to note that the SCO ID and file path in here will likely not be the same as the original file you uploaded to your room, it's a new SCO id (I believe SCOs of this type are called referenced scos) and new path.
    Now...if I was going to build some sort of player which would play all these files in one screen to make a recording...I might not want to use Connect's code here. If you know the file path to the shared content (from FTContent), you could easily view it with the content URL (conveniently also in FtContent). I'm not a coder, but I'm envisioning something like Presenter's GUI where you've got the presentation's content in the main area, and a video file (if there is one) playing back on the side.
    Anyways, food for thought if you want to try to go about this. Connect recordings are incredibly complex and they come with a big learning curve, but if you can make sense of them the knowledge is quite valuable.

  • What would make my Cp7 course get hung up during the loading process when launching on LMS?

    In Cp7, I created a SCORM project to post on our LMS and submitted it for testing. Our LMS department confirmed it worked successfully, and in fact I tested it myself – it worked great. However, my work team asked me to make some changes to the content before pushing it out of the testing phase to launch company-wide. Per company policy, making changes to a course means resubmitting it once more to be tested before making it available to everybody. So I made the requested changes and resubmitted it for testing. Now our LMS department reports the course will not launch properly – it gets stuck in the loading process. It shows “Loading...” endlessly but never loads.
    When I submitted the updated version for testing, I kept all the project settings that worked successfully the first time. The changes I made to the project were:
    I added a slide toward the beginning (slide 2) to give the user navigation tips.
    I set slides so that each one must play all the way through before the user can proceed to the next slide (I think I did this simply by removing “play” on the playbar”).
    Under table of contents settings, I checked “navigate visited slides only,” so the user can navigate backwards using the contents bar at left, but can only navigate forward to slides that have already played or to the next slide in the queue.
    I broke up a couple of lengthy multiple choice questions into shorter ones (for an additional two slides).
    Is there any reason one of these changes would make the course get hung up during loading?
    Is there a size limit Cp7 projects which, if exceeded, might be causing such an issue? 
    Or does anyone have ideas about what else might be making it get hung up in the loading process?
    Thank you, any and all, for your feedback.

    back up the iPhoto library like any other backup - make a copy of the iPhoto library in case of problems
    you Depress the option (alt) and command keys and launch iPhoto - anyplace you can launch iPhoto you do this - keep the keys down until you get the rebuild window
    LN

  • Some questions about the integration between BIEE and EBS

    Hi, dear,
    I'm a new bie of BIEE. In these days, have a look about BIEE architecture and the BIEE components. In the next project, there are some work about BIEE development based on EBS application. I have some questions about the integration :
    1) generally, is the BIEE database and application server decentralized with EBS database and application? Both BIEE 10g and 11g version can be integrated with EBS R12?
    2) In BIEE administrator tool, the first step is to create physical tables. if the source appliation is EBS, is it still needed to create the physical tables?
    3) if the physical tables creation is needed, how to complete the data transfer from the EBS source tables to BIEE physical tables? which ETL tool is prefer for most developers? warehouse builder or Oracle Data Integration?
    4) During data transfer phase, if there are many many large volume data needs to transfer, how to keep the completeness? for example, it needs to transfer 1 million rows from source database to BIEE physical tables, when 50%is completed, the users try to open the BIEE report, can they see the new 50% data on the reports? is there some transaction control in ETL phase?
    could anyone give some guide for me? I'm very appreciated if you can also give any other information.
    Thanks in advance.

    1) generally, is the BIEE database and application server decentralized with EBS database and application? Both BIEE 10g and 11g version can be integrated with EBS R12?You, shud consider OBI Application here which uses OBIEE as a reporting tool with different pre-built modules. Both 10g & 11g comes with different versions of BI apps which supports sources like Siebel CRM, EBS, Peoplesoft, JD Edwards etc..
    2) In BIEE administrator tool, the first step is to create physical tables. if the source appliation is EBS, is it still needed to create the physical tables?Its independent of any soure. This is OBIEE modeling to create RPD with all the layers. If you build it from scratch then you will require to create all the layers else if BI Apps is used then you will get pre-built RPD along with other pre-built components.
    3) if the physical tables creation is needed, how to complete the data transfer from the EBS source tables to BIEE physical tables? which ETL tool is prefer for most developers? warehouse builder or Oracle Data Integration?BI apps comes with pre-built ETL mapping to use with the tools majorly with Informatica. Only BI Apps 7.9.5.2 comes with ODI but oracle has plans to have only ODI for any further releases.
    4) During data transfer phase, if there are many many large volume data needs to transfer, how to keep the completeness? for example, it needs to transfer 1 million rows from source database to BIEE physical tables, when 50%is completed, the users try to open the BIEE report, can they see the new 50% data on the reports? is there some transaction control in ETL phase?User will still see old data because its good to turn on Cache and purge it after every load.
    Refer..http://www.oracle.com/us/solutions/ent-performance-bi/bi-applications-066544.html
    and many more docs on google
    Hope this helps

  • Question about sql batch process in java app

    hi all
    i have few questions about using batch process in the java.sql package. the addBatch method can take sql statements like inserts or updates. can we use a mixture of insert and update then? can we use prepared statement for this? it's just for performance consideration. thanks in advance.

    hi all
    i have few questions about using batch process in the
    java.sql package. the addBatch method can take sql
    statements like inserts or updates. addBatch() is a method that has no parameters. It doesn't 'add' sql statements.
    can we use a
    mixture of insert and update then? can we use
    prepared statement for this? it's just for
    performance consideration. thanks in advance.The point of batching is that you take something that is invariant and then 'add' a variant part.
    Thus a single insert has an invariant part (the table and specific columns) and a variant part (the data for each row by column.)
    You can use anything that is valid SQL (for jdbc, driver, database) and use it presuming your database allows that particular usage in batching. But that does require some regular pattern - it won't work if your usage is random. Nor will it work if some statements need to be executed only some of the time. Finally also note that transaction processing will often require smaller chunks - you can't insert a million rows in one batch.

  • Question about the CSS behavior when using layer 3 sticky and sticky table

    Hi everyone,
    I have a question about the CSS behavior when using layer 3 sticky and sticky table is full.
    If I configure layer 3 sticky and specify the inactivity timeout as below, how does the CSS
    handle subsequent needed sticky requests ?
    advanced-balance sticky-srcip
    sticky-inact-timeout 30
    CSS document says that
    Note:
    If you use the sticky-inact-timeout command to specify the inactivity timeout
    period on a sticky connection, when the sticky table becomes full and none of
    the entries have expired from the sticky table, the CSS rejects subsequent
    needed sticky requests.
    My question is what is the next reaction by doing the CSS if the CSS is in the
    following condition:
    when the sticky table becomes full and none of the entries have expired from
    the sticky table, the CSS rejects subsequent needed sticky requests
    Does CSS just rejects/drops subsequent needed sticky requests ?
    or
    Does CSS does not stick subsequence requests to particular service but CSS forward
    subsequence requests with round-robin basis ? which means if the sticky table is full,
    the CSS just works round-robin load balancing fashion for subsequence requests ?
    Your information would be appreciated.
    Best regards,

    Hello,
    There is a good document explaining this on Cisco web site
    http://www.cisco.com/en/US/products/hw/contnetw/ps789/products_tech_note09186a0080094b4b.shtml
    It depends if the sticky-inact-timeout is used or not. If not, it's FIFO (the oldest entry in the sticky table is removed). If yes, the CSS will reject the next sticky request.
    Rgds,
    Gaetan
    Rgds
    Gaetan

  • Question about Finder-Load-Beans flag

    Hi all,
    I've read that the Finder-Load-Beans flag could yield some valuable gains in performance
    but:
    1) why is it suggested to do individual gets of methods within the same Transaction
    ? (tx-Required).
    2) this strategy is useful only for small sets of data, isn't it? I imagine I
    would choose Finder-Load-Beans to false (or JDBC) for larger sets of data.
    3) A last question: its default value is true or false ?
    Thanks
    Francesco

    Because if there are different transactions where the get method is called
    then the state/data of the bean would most be reloaded from the database. A
    new transactions causes the ejbLoad method to be invoked in the beginning
    and the ejbStore at the end. That is the usual case but there are other ways
    to modify this behavior.
    Thanks
    Gaurav
    "Francesco" <[email protected]> wrote in message
    news:[email protected]...
    >
    Hi thorick,
    I have found this in the newsgroup. It's from R.Woolen answering
    a question about Finder-Load-Beans flag.
    "Consider this case:
    tx.begin();
    Collection c = findAllEmployeesNamed("Rob");
    Iterator it = c.iterator();
    while (it.hasNext()) {
    Employee e = (Employee) it.next(); System.out.println("Favorite color is:"+ e.getFavColor());
    tx.commit();
    With CMP (and finders-load-beans set to its default true value), thefindAllEmployeesNamed
    finder will load all the employees with the name of rob. The getFavColormethods
    do not hit the db because they are in the same tx, and the beans arealready loaded
    in the cache.
    It's the big CMP performance advantage."
    So I wonder why this performance gain can be achieved when the iterationis inside
    a transaction.
    Thanks
    regards
    Francesco
    thorick <[email protected]> wrote:
    1) why is it suggested to do individual gets of methods within thesame Transaction
    ? (tx-Required).I'm not sure about the context of this question (in what document,
    paragraph
    is this
    mentioned).
    2) this strategy is useful only for small sets of data, isn't it? Iimagine I
    would choose Finder-Load-Beans to false (or JDBC) for larger sets ofdata.
    >
    If you know that you will be accessing the fields of all the Beans that
    you get back from a
    finder,
    then you will realize a significant performance gain. If one selects
    100s or more beans
    using
    a finder, but only accesses the fields for a few, then there may be some
    performance cost.
    It could
    depend on how large some of the fields are. I'd guess that the cost
    of 1 hit to the DB per
    bean vs.
    the cost of 1 + maybe 1 more hit to the DB per bean, would usually be
    less. A performance
    test using
    your actual apps beans would be the only way to know for sure.
    3) A last question: its default value is true or false ?The default is 'True'
    -thorick

  • Question about the custom panel language

    I have a question about the custom panel language...
    The document you provide seems to lack details on some features. Namely the icon and picture widgets. I see from looking at the examples and other vendor's web pages that these features exist, but I don't find any detailed descriptions of them in the documentation. Is there a more complete document describing these and other features...
    http://www.adobe.com/products/xmp/custompanel.html
    Alternatively, can someone fill me in on the syntax and options for at least the icon and picture widget. For instance, how do you load external icons or pictures...
    Tom

    Gunar,
    It could be interesting to have something like
    icon(url: 'http://www.adobe.com/Images/logo.gif', width: 20, height: 20);
    or better
    picture(url: 'http://www.adobe.com/Images/logo.gif', width: 20, height: 20);
    for the pictures and
    include(url: 'http://www.adobe.com/xml/custompanel/camera1.txt');
    for include the cusmtom panel's dynamic portions
    Juan Pablo

  • Theoretical question about the monitor

    Hello,
    in my diploma thesis I compare the bw with a reference modell to check wether the bw is a "real" dwh or not.
    The reference describes the monitor as component which observes the source systems and checks for changes to afford incremential loading.
    The monitor of the bw observes the loading processes in the bw.
    So there's a difference. Nontheless there must be further monitor which is conform to the reference, or? I did not find any information about that one.
    Does anyone know about it?
    Tobias

    Hi Tobias,
    as you already know, BW monitor it's not a delta monitor (as your reference describes), but only an administration tool that allows to follow your dataloading...
    Generally speaking, in BW there is no a specific component "which observes the source systems and checks for changes to afford incremential loading" because (but only for delta-capable datasource) this is done in source system itself (and this is the integration concept typical of all SAP products) thanks to different mechanism (timestamp field for some flows, delta queue for others, generic delta and so on)...
    Anyway, even BW has a delta management mechanism, but related to its datamart (loading between an ODS and a cube, for example) only and you can observe it from the changelog table and thanks to ROCANCEL field...
    Hope it helps!
    Bye,
    Roberto

Maybe you are looking for