Research

NAVIGATION
CATEGORIES
REFERRENCE
LINKS
  • Semaphore/Mutex interview question

    12 answers - 1357 bytes - related search similar search Add To My Delicious Add To My Stumble Upon Add To My Google Mark Add To My Facebook Add To My Digg Add To My Reddit

    So I had a phone interivew for an embedded software position with a
    major communications company and was asked some questions that I didn't
    seem to answer to the interviewer's satisfaction. enough, I got
    called back for an on-site interview. Anyways, I want to run these
    questions by others to see what the real answer is.
    1) What is the difference between a mutex and a binary semaphore?
    I couldn't really think of a difference but he seemed to be leading me
    to an answer. He asked if you can use a mutex in an interrupt service
    routine and then asked if you could use a semaphore within an interrupt
    service routine. of the questions he asked during the interview was
    whether I had heard the term "munching". Well, after the interview I
    looked it up and it is some term specific to VxWorks. Not having worked
    with VxWorks before I decided to read a VxWorks manual online. It seems
    like almost all his questions were related to VxWorks even though he
    didn't say so.
    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.
    Thanks.
  • No.1 | | 1791 bytes | |

    Fri, 15 Sep 2006 05:49:51 +0200 schreef joshc <josh.curtz@gmail.com>:
    So I had a phone interivew for an embedded software position with a
    major communications company and was asked some questions that I didn't
    seem to answer to the interviewer's satisfaction. enough, I got
    called back for an on-site interview. Anyways, I want to run these
    questions by others to see what the real answer is.

    1) What is the difference between a mutex and a binary semaphore?

    I guess it is valid to say that because these clumsy primitives are error
    prone and cause unwanted effects, the answer to this question is
    irrelevant. But I guess that I am spoilt for knowing easier-to-understand
    and less ugly ways to make processes communicate.

    I couldn't really think of a difference but he seemed to be leading me
    to an answer. He asked if you can use a mutex in an interrupt service
    routine and then asked if you could use a semaphore within an interrupt
    service routine. of the questions he asked during the interview was
    whether I had heard the term "munching". Well, after the interview I
    looked it up and it is some term specific to VxWorks. Not having worked
    with VxWorks before I decided to read a VxWorks manual online. It seems
    like almost all his questions were related to VxWorks even though he
    didn't say so.

    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.

    The question indeed relies on the interpretations of these concepts by a
    specific implementor.
  • No.2 | | 1714 bytes | |


    joshc wrote:
    So I had a phone interivew for an embedded software position with a
    major communications company and was asked some questions that I didn't
    seem to answer to the interviewer's satisfaction. enough, I got
    called back for an on-site interview. Anyways, I want to run these
    questions by others to see what the real answer is.

    1) What is the difference between a mutex and a binary semaphore?

    The purpose differs. Expand the name MUTEX is MUTual EXclusion. You use
    it to prevent two or more processes from executing the same section of
    code at the same time. Whereas a semaphore is just a signalling
    mechanism. Why he would specifically use a binary semaphore in the
    questions if strange.

    I couldn't really think of a difference but he seemed to be leading me
    to an answer. He asked if you can use a mutex in an interrupt service
    routine and then asked if you could use a semaphore within an interrupt
    service routine. of the questions he asked during the interview was
    whether I had heard the term "munching". Well, after the interview I
    looked it up and it is some term specific to VxWorks. Not having worked
    with VxWorks before I decided to read a VxWorks manual online. It seems
    like almost all his questions were related to VxWorks even though he
    didn't say so.

    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.

    Thanks.

    hth,
    ed

  • No.3 | | 1070 bytes | |


    Ed Prochak wrote:
    joshc wrote:
    So I had a phone interivew for an embedded software position with a
    major communications company and was asked some questions that I didn't
    seem to answer to the interviewer's satisfaction. enough, I got
    called back for an on-site interview. Anyways, I want to run these
    questions by others to see what the real answer is.

    1) What is the difference between a mutex and a binary semaphore?

    The purpose differs. Expand the name MUTEX is MUTual EXclusion. You use
    it to prevent two or more processes from executing the same section of
    code at the same time. Whereas a semaphore is just a signalling
    mechanism. Why he would specifically use a binary semaphore in the
    questions if strange.

    hth,
    ed

    Yes, but a semaphore can be used for mutual exclusion too. The
    differences I found in VxWorks' implementation of mutexes and
    semaphores are very implementation specific. Any more ideas on common
    differences between mutexas and semaphores?

  • No.4 | | 1872 bytes | |

    Maybe the difference is in how the waitng processes are woken up. Maybe
    relating to which process and structures are responsible for managing each.
    I would think that semaphores are individually managed by each owning
    process while mutexes are centralling managed. This difference may make a
    big difference with respect to how interrupts impact each.

    Probably not correct, but my 2 cents.

    Hicks

    "joshc" <josh.curtz@gmail.comwrote in message
    news:1158292191.257574.114150@
    So I had a phone interivew for an embedded software position with a
    major communications company and was asked some questions that I didn't
    seem to answer to the interviewer's satisfaction. enough, I got
    called back for an on-site interview. Anyways, I want to run these
    questions by others to see what the real answer is.

    1) What is the difference between a mutex and a binary semaphore?
    I couldn't really think of a difference but he seemed to be leading me
    to an answer. He asked if you can use a mutex in an interrupt service
    routine and then asked if you could use a semaphore within an interrupt
    service routine. of the questions he asked during the interview was
    whether I had heard the term "munching". Well, after the interview I
    looked it up and it is some term specific to VxWorks. Not having worked
    with VxWorks before I decided to read a VxWorks manual online. It seems
    like almost all his questions were related to VxWorks even though he
    didn't say so.

    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.

    Thanks.

  • No.5 | | 1602 bytes | |

    joshc wrote:
    So I had a phone interivew for an embedded software position with a
    major communications company and was asked some questions that I didn't
    seem to answer to the interviewer's satisfaction. enough, I got
    called back for an on-site interview. Anyways, I want to run these
    questions by others to see what the real answer is.

    1) What is the difference between a mutex and a binary semaphore?
    I couldn't really think of a difference but he seemed to be leading me
    to an answer. He asked if you can use a mutex in an interrupt service
    routine and then asked if you could use a semaphore within an interrupt
    service routine. of the questions he asked during the interview was
    whether I had heard the term "munching". Well, after the interview I
    looked it up and it is some term specific to VxWorks. Not having worked
    with VxWorks before I decided to read a VxWorks manual online. It seems
    like almost all his questions were related to VxWorks even though he
    didn't say so.

    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.

    A mutex is used to ensure that only a single "thread" accesses
    a resource. A (binary) semaphore is a mechanism that can be
    used to implement a mutex. Not all semaphores are mutex'es
    (just like not all counters are semaphores).
  • No.6 | | 1597 bytes | |


    "joshc" <josh.curtz@gmail.comwrote in message
    news:1158292191.257574.114150@
    So I had a phone interivew for an embedded software position with a
    major communications company and was asked some questions that I didn't
    seem to answer to the interviewer's satisfaction. enough, I got
    called back for an on-site interview. Anyways, I want to run these
    questions by others to see what the real answer is.

    1) What is the difference between a mutex and a binary semaphore?
    I couldn't really think of a difference but he seemed to be leading me
    to an answer. He asked if you can use a mutex in an interrupt service
    routine and then asked if you could use a semaphore within an interrupt
    service routine. of the questions he asked during the interview was
    whether I had heard the term "munching". Well, after the interview I
    looked it up and it is some term specific to VxWorks. Not having worked
    with VxWorks before I decided to read a VxWorks manual online. It seems
    like almost all his questions were related to VxWorks even though he
    didn't say so

    In VxWorks perspective, a mutex is roughly a semaphore with an initial value
    of 1.
    (With a few restrictions and variations).

    Steve

    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.

    Thanks.

  • No.7 | | 389 bytes | |

    joshc wrote:

    1) What is the difference between a mutex and a binary semaphore?

    Thanks.

    Not sure for all RTS-es, but usually a mutex has priority inheritance,
    while a semaphore has not.
    Then a mutex is usually used for mutual exclusion of critical paths,
    while a semaphore is used for inter-thread synchronization and event
    notifications.

  • No.8 | | 692 bytes | |

    Adrian wrote:
    joshc wrote:

    1) What is the difference between a mutex and a binary semaphore?

    Thanks.

    Not sure for all RTS-es, but usually a mutex has priority inheritance,
    while a semaphore has not.
    Then a mutex is usually used for mutual exclusion of critical paths,
    while a semaphore is used for inter-thread synchronization and event
    notifications.

    To be more specific, if you have two concurrent threads, with mutexes
    you will always have a pair of get/put mutex in each thread.
    With semaphores, you may wait (get) for the semaphore in one thread and
    put the semaphore in the other thread. Never use this with mutexes

  • No.9 | | 3014 bytes | |

    In article <1158292191.257574.114150@h48g2000cwc.googlegroups. com>,
    joshc <josh.curtz@gmail.comwrote:
    >So I had a phone interivew for an embedded software position with a
    >major communications company and was asked some questions that I didn't
    >seem to answer to the interviewer's satisfaction. enough, I got
    >called back for an on-site interview. Anyways, I want to run these
    >questions by others to see what the real answer is.


    You've gotten a number of slightly different opinions, which should be
    a strong hint that there's no True Definition

    >1) What is the difference between a mutex and a binary semaphore?


    The most obvious answer that pops into my head comes from PSIX: A
    mutex has an owner, a semaphore (binary or otherwise) does not. Any
    old thread can post a semaphore, but only the owner of a mutex can
    unlock it.

    of the questions he asked during the interview was
    >whether I had heard the term "munching". Well, after the interview I
    >looked it up and it is some term specific to VxWorks. Not having worked
    >with VxWorks before I decided to read a VxWorks manual online. It seems
    >like almost all his questions were related to VxWorks even though he
    >didn't say so.


    That's a new term by me (and an amusing way to add some functionality
    that simply belongs hidden in the c++ driver, now that I've looked it
    up). But then, I've done hardly any work with vxWorks, preferring
    "real"[1] operating systems (with memory protection and process models)
    for larger projects, and much lighter weight executives for smaller
    projects.

    >Is there really one "correct" answer to the mutex vs binary sempahore
    >question? There definitely does seem to be a difference in VxWorks
    >as mutexes handle priority inversion. It seems to be S / library
    >dependent and thus seems a bit unfair, but correct me if I'm wrong.


    The reason a mutex *can* handle priority inversion is that there's a
    known owner to boost the priority of when the time comes.

    I suspect where the interviewer was leading with the "which one can you
    use in an interrupt handler" is deeper into the vxWorks-only space where
    an interrupt handler is running in the task environment of whatever
    thread it happened to interrupt. That makes the meaning of owning a
    mutex inside an interrupt handler somewhat problematic.

    quite common way of communicating out of an interrupt handler
    into task space is to have the interrupt handler signal a semaphore,
    which the task that's waiting for the interrupt is blocked on. The S
    will then make that task ready, and when time comes for a scheduling
    decision, it might even get the/a CPU.

    [1] K, that's flame bait. My bias showing again
  • No.10 | | 925 bytes | |

    joshc schrieb:

    routine and then asked if you could use a semaphore within an interrupt
    service routine. of the questions he asked during the interview was
    whether I had heard the term "munching". Well, after the interview I
    looked it up and it is some term specific to VxWorks. Not having worked
    with VxWorks before I decided to read a VxWorks manual online. It seems
    like almost all his questions were related to VxWorks even though he
    didn't say so.

    In the interrupt you should not wait a resource so you should only set
    the semaphore and never wait

    Adib.
    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.

    Thanks.

  • No.11 | | 790 bytes | |

    Quote:
    Originally Posted by joshc
    Is there really one "correct" answer to the mutex vs binary sempahore
    question? There definitely does seem to be a difference in VxWorks as
    mutexes handle priority inversion. It seems to be S / library
    dependent and thus seems a bit unfair, but correct me if I'm wrong.
    Thanks.


    It's really this simple. The task that owns mutex semaphore can take it again, while if it attempts to take again the binary semaphore that it owns, it'll fail. In other words, vxworks mutex sema are recursive (re-entrant), while binary are not.

    Yes, I have access to vxworks source code, so I looked it up.
  • No.12 | | 651 bytes | |

    This is the most dumbest question ever asked in interviews.

    "What is the diff between a mutex and a semaphore".

    Well the answer is:

    "which context are ya referring to ? Linux, VxWorks, Symbian, pThreads, Windows ? "

    Caus these words have diff connotations in diff Oses.

    Mutexes and binary semaphores are the same in Linux.
    Binary semaphores don't exist in pThreads, hence mutexes and semaphores (which the interviewer means to say counting semaphores) are different in pThreads.
    As already mentioned in this thread there are subtle differences in vxWorks as well.

    Thanks,
    Timothy Noel

Re: Semaphore/Mutex interview question


max 4000 letters.
Your nickname that display:
In order to stop the spam: 8 + 7 =
QUESTION ON "Research"

EMSDN.COM