DPDK patches and discussions
 help / color / mirror / Atom feed
From: Santosh Shukla <sshukla@mvista.com>
To: David Marchand <david.marchand@6wind.com>
Cc: "dev@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH v7 4/4] eal/linux: vfio: add pci ioport support
Date: Mon, 8 Feb 2016 15:10:09 +0530	[thread overview]
Message-ID: <CAAyOgsYncbxPCcW9-Y+KX9NQNVzS6f7V47j4itvnYNApoc320w@mail.gmail.com> (raw)
In-Reply-To: <CALwxeUvzym044mw-2hodb-BhzHwrtOy_axZi3fagVuXADec6Jw@mail.gmail.com>

On Mon, Feb 8, 2016 at 2:21 PM, David Marchand <david.marchand@6wind.com> wrote:
> On Sun, Feb 7, 2016 at 2:51 PM, Santosh Shukla <sshukla@mvista.com> wrote:
>> @@ -999,37 +1000,56 @@ int
>>  pci_vfio_ioport_map(struct rte_pci_device *dev, int bar,
>>                     struct rte_pci_ioport *p)
>
> p is passed as a value, not a reference ...
>
>>  {
>> -       RTE_SET_USED(dev);
>> -       RTE_SET_USED(bar);
>> -       RTE_SET_USED(p);
>> -       return -1;
>> +       if (bar < VFIO_PCI_BAR0_REGION_INDEX ||
>> +           bar > VFIO_PCI_BAR5_REGION_INDEX) {
>> +               RTE_LOG(ERR, EAL, "invalid bar (%d)!\n", bar);
>> +               return -1;
>> +       }
>> +
>> +       p = rte_zmalloc("VFIO_IOPORT", sizeof(*p), 0);
>
> ... so I don't think this allocation does what you expected.
>
> Anyway, you don't need to allocate a rte_pci_ioport object with current api.
> You already have a valid object passed by caller.
> You only need to initialise it.
>
>
>> +       if (p == NULL) {
>> +               RTE_LOG(ERR, EAL, "cannot alloc vfio ioport mem\n");
>> +               return -1;
>> +       }
>> +
>> +       p->dev = dev;
>
> Does not hurt to do this, but p->dev is already set by caller on ret
> == 0 (rte_eal_pci_ioport_map).
>
>

so far I didn't found caller setting p->dev (looked at virtio pmd
driver), In-fact I tried to use w/o setting p->dev=dev and application
crashed, stating segfault :).

So I am keeping this one in next revision.
>>
>>  void
>>  pci_vfio_ioport_read(struct rte_pci_ioport *p,
>>                      void *data, size_t len, off_t offset)
>>  {
>> -       RTE_SET_USED(p);
>> -       RTE_SET_USED(data);
>> -       RTE_SET_USED(len);
>> -       RTE_SET_USED(offset);
>> +       const struct rte_intr_handle *intr_handle = &p->dev->intr_handle;
>
> Missing blank line between declaration and code.

checkpatch.sh didn't ntocied though, which all param you use in
checkpatch, I just do
./checkpatch.sh --no-tree *.patch

>
>> +       if (pread64(intr_handle->vfio_dev_fd, data,
>> +                   len, p->offset + offset) <= 0)
>> +               RTE_LOG(ERR, EAL,
>> +                       "Can't read from PCI bar (%" PRIu64 ") : offset (%x)\n",
>> +                       VFIO_GET_REGION_IDX(p->offset), (int)offset);
>>  }
>>
>>  void
>>  pci_vfio_ioport_write(struct rte_pci_ioport *p,
>>                       const void *data, size_t len, off_t offset)
>>  {
>> -       RTE_SET_USED(p);
>> -       RTE_SET_USED(data);
>> -       RTE_SET_USED(len);
>> -       RTE_SET_USED(offset);
>> +       const struct rte_intr_handle *intr_handle = &p->dev->intr_handle;
>
> Idem.
>
>> +       if (pwrite64(intr_handle->vfio_dev_fd, data,
>> +                    len, p->offset + offset) <= 0)
>> +               RTE_LOG(ERR, EAL,
>> +                       "Can't write to PCI bar (%" PRIu64 ") : offset (%x)\n",
>> +                       VFIO_GET_REGION_IDX(p->offset), (int)offset);
>>  }
>>
>>  int
>>  pci_vfio_ioport_unmap(struct rte_pci_ioport *p)
>>  {
>> -       RTE_SET_USED(p);
>> -       return -1;
>> +       if (p == NULL)
>> +               return -1;
>> +       else {
>> +               rte_free(p);
>> +               return 0;
>> +       }
>>  }
>
> Since you have nothing to allocate, nothing to free here ?
>

Removed, Thanks
>
> --
> David Marchand

  reply	other threads:[~2016-02-08  9:40 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-02-07 13:51 [dpdk-dev] [PATCH v7 0/4] Add virtio support for arm/arm64 Santosh Shukla
2016-02-07 13:51 ` [dpdk-dev] [PATCH v7 1/4] eal/linux: never check iopl for arm Santosh Shukla
2016-02-18  5:26   ` Santosh Shukla
2016-02-07 13:51 ` [dpdk-dev] [PATCH v7 2/4] virtio: Introduce config RTE_VIRTIO_INC_VECTOR Santosh Shukla
2016-02-07 21:25   ` Thomas Monjalon
2016-02-08  5:45     ` Santosh Shukla
     [not found]       ` <CAAyOgsZO+6+kFZZZM203fPR3AmVYB0v7j3-f+DawZOCuR-AVvQ@mail.gmail.com>
     [not found]         ` <20160215105743.GB21426@yliu-dev.sh.intel.com>
     [not found]           ` <CAAyOgsaT5TcsPfum8x6yzAJAz=5N+c5QebEn7KCyJn7oK=VMsw@mail.gmail.com>
2016-02-16  3:05             ` Yuanhan Liu
2016-02-19  4:46               ` Santosh Shukla
2016-02-19  6:42                 ` Yuanhan Liu
2016-02-22  2:03                   ` Xie, Huawei
2016-02-22  4:14                     ` Santosh Shukla
2016-02-22 10:22                       ` Thomas Monjalon
2016-02-07 13:51 ` [dpdk-dev] [PATCH v7 3/4] eal/linux: vfio: ignore mapping for ioport region Santosh Shukla
2016-02-08  9:15   ` Burakov, Anatoly
2016-02-18  5:26     ` Santosh Shukla
2016-02-07 13:51 ` [dpdk-dev] [PATCH v7 4/4] eal/linux: vfio: add pci ioport support Santosh Shukla
2016-02-08  8:51   ` David Marchand
2016-02-08  9:40     ` Santosh Shukla [this message]
2016-02-21 14:17 ` [dpdk-dev] [PATCH v9 0/3] Add virtio support for arm/arm64 Santosh Shukla
2016-02-21 14:17   ` [dpdk-dev] [PATCH v9 1/3] eal/linux: never check iopl for arm Santosh Shukla
2016-02-21 14:18   ` [dpdk-dev] [PATCH v9 2/3] eal/linux: vfio: ignore mapping for ioport region Santosh Shukla
2016-02-21 14:18   ` [dpdk-dev] [PATCH v9 3/3] eal/linux: vfio: add pci ioport support Santosh Shukla
2016-02-22  5:41   ` [dpdk-dev] [PATCH v9 0/3] Add virtio support for arm/arm64 Yuanhan Liu
2016-02-23  6:11     ` Santosh Shukla
2016-02-24 10:45     ` Thomas Monjalon

Reply instructions:

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

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

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

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

  git send-email \
    --in-reply-to=CAAyOgsYncbxPCcW9-Y+KX9NQNVzS6f7V47j4itvnYNApoc320w@mail.gmail.com \
    --to=sshukla@mvista.com \
    --cc=david.marchand@6wind.com \
    --cc=dev@dpdk.org \
    /path/to/YOUR_REPLY

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

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).