qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Cao jin <caoj.fnst@cn.fujitsu.com>
Cc: qemu-devel@nongnu.org, stefano.stabellini@eu.citrix.com
Subject: Re: [Qemu-devel] [PATCH v4] igd-passthrough-i440FX: convert to realize()
Date: Tue, 29 Dec 2015 13:25:50 +0200	[thread overview]
Message-ID: <20151229130853-mutt-send-email-mst@redhat.com> (raw)
In-Reply-To: <1451378763-25398-1-git-send-email-caoj.fnst@cn.fujitsu.com>

On Tue, Dec 29, 2015 at 04:46:03PM +0800, Cao jin wrote:
> Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com>
> ---
> changelog v4:
> 1. strip the bugfix code, but I guess this patch should be applied after
>    that bugfix patch.
> 2. Other little fix as per previous review
> 
> Some explanation to previous review:
> 1. error_setg_errno() I use in this patch will print the strerror(errno), so I
>    guess it will be informative enough? example:
> 
>    error_setg_errno(errp, errno, "open %s fail", path); will print:
>    "open bluhbluh fail: string_from_strerrer(errno)"

I'd prefer some mention of why this is attempted too.
E.g. "igd passthrough : open bluhbluh fail"

> 2. snprintf() has tiny tiny chance to fail,

It just formats a string into a buffer. the buffer
is pre-allocated, and we know it's large enough for the data.
How can it fail?


> I don`t assert here, because this
>    device is on-board, it init during machine initialization, in case it fails,
>    the errp we set will results in error_report(err) and exit(1), at least can
>    let user know why fail. see the call-chain:
> 
>    pc_xen_hvm_init_pci->pc_init1->i440fx_init->pci_create_simple-> pci_create_simple_multifunction->qdev_init_nofail

unlike open failing, there's nothing user can do, this means there's
a coding bug somewhere, and assert is easier to debug.

> About test:
> 1. Compiled
> 2. Did a dirty hack to force create/realize this device in pc_init1(), prove
>    that the realizing process is ok, I will reply this mail later to attch the
>    screenshot evidence

Good job, thanks!
You should also Cc Tiejun Chen <tiejun.chen@intel.com> who wrote this
code, presumably with access to hardware to test on.


>  hw/pci-host/piix.c | 30 +++++++++++++++---------------
>  1 file changed, 15 insertions(+), 15 deletions(-)
> 
> diff --git a/hw/pci-host/piix.c b/hw/pci-host/piix.c
> index a9cb983..e91570f 100644
> --- a/hw/pci-host/piix.c
> +++ b/hw/pci-host/piix.c
> @@ -761,7 +761,7 @@ static const IGDHostInfo igd_host_bridge_infos[] = {
>      {0xa8, 4},  /* SNB: base of GTT stolen memory */
>  };
>  
> -static int host_pci_config_read(int pos, int len, uint32_t *val)
> +static void host_pci_config_read(int pos, int len, uint32_t *val, Error **errp)
>  {
>      char path[PATH_MAX];
>      int config_fd;
> @@ -769,19 +769,19 @@ static int host_pci_config_read(int pos, int len, uint32_t *val)
>      /* Access real host bridge. */
>      int rc = snprintf(path, size, "/sys/bus/pci/devices/%04x:%02x:%02x.%d/%s",
>                        0, 0, 0, 0, "config");
> -    int ret = 0;
>  
>      if (rc >= size || rc < 0) {
> -        return -ENODEV;
> +        error_setg_errno(errp, errno, "snprintf err");
>      }
>  
>      config_fd = open(path, O_RDWR);
>      if (config_fd < 0) {
> -        return -ENODEV;
> +        error_setg_errno(errp, errno, "open %s fail", path);
> +        return;
>      }
>  
>      if (lseek(config_fd, pos, SEEK_SET) != pos) {
> -        ret = -errno;
> +        error_setg_errno(errp, errno, "lseek err");
>          goto out;
>      }
>  
> @@ -789,32 +789,32 @@ static int host_pci_config_read(int pos, int len, uint32_t *val)
>          rc = read(config_fd, (uint8_t *)val, len);
>      } while (rc < 0 && (errno == EINTR || errno == EAGAIN));
>      if (rc != len) {
> -        ret = -errno;
> +        error_setg_errno(errp, errno, "read err");
>      }
>  
>  out:
>      close(config_fd);
> -    return ret;
>  }
>  
> -static int igd_pt_i440fx_initfn(struct PCIDevice *pci_dev)
> +static void igd_pt_i440fx_realize(struct PCIDevice *pci_dev, Error **errp)
>  {
>      uint32_t val = 0;
> -    int rc, i, num;
> +    int i, num;
>      int pos, len;
> +    Error *local_err = NULL;
>  
>      num = ARRAY_SIZE(igd_host_bridge_infos);
>      for (i = 0; i < num; i++) {
>          pos = igd_host_bridge_infos[i].offset;
>          len = igd_host_bridge_infos[i].len;
> -        rc = host_pci_config_read(pos, len, &val);
> -        if (rc) {
> -            return -ENODEV;
> +
> +        host_pci_config_read(pos, len, &val, &local_err);
> +        if (local_err) {
> +            error_propagate(errp, local_err);
> +            return;
>          }
>          pci_default_write_config(pci_dev, pos, cpu_to_le32(val), len);
>      }
> -
> -    return 0;
>  }
>  
>  static void igd_passthrough_i440fx_class_init(ObjectClass *klass, void *data)
> @@ -822,7 +822,7 @@ static void igd_passthrough_i440fx_class_init(ObjectClass *klass, void *data)
>      DeviceClass *dc = DEVICE_CLASS(klass);
>      PCIDeviceClass *k = PCI_DEVICE_CLASS(klass);
>  
> -    k->init = igd_pt_i440fx_initfn;
> +    k->realize = igd_pt_i440fx_realize;
>      dc->desc = "IGD Passthrough Host bridge";
>  }
>  
> -- 
> 2.1.0
> 
> 

  parent reply	other threads:[~2015-12-29 11:25 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-12-29  8:46 [Qemu-devel] [PATCH v4] igd-passthrough-i440FX: convert to realize() Cao jin
2015-12-29  9:25 ` Cao jin
2015-12-29 11:25 ` Michael S. Tsirkin [this message]
2015-12-29 13:27   ` Cao jin
2016-01-04 14:47     ` Stefano Stabellini
2016-01-04 15:33       ` Lars Kurth
2016-01-04 15:41         ` Stefano Stabellini
2016-01-06 10:21           ` [Qemu-devel] [Xen-devel] " Lars Kurth
2016-01-06 12:18             ` Stefano Stabellini
2016-01-07  1:32               ` Hao, Xudong
2016-01-08 11:57                 ` Stefano Stabellini
2016-01-11  1:46                   ` Hao, Xudong
2016-01-11  9:53                   ` Hao, Xudong
2016-01-11 10:32                     ` Gerd Hoffmann
2016-01-12  8:41                       ` Hao, Xudong
2016-01-12  8:47                         ` Michael S. Tsirkin
2016-01-12  9:50                           ` Hao, Xudong
2016-01-12 10:24                             ` Gerd Hoffmann
2016-01-13  1:55                               ` Hao, Xudong
2016-01-13  7:37                               ` Hao, Xudong
2016-01-11 10:46                     ` Stefano Stabellini
2016-01-11 11:01                       ` Michael S. Tsirkin
2016-01-12  9:54                       ` Hao, Xudong
2016-01-12  2:01                     ` Cao jin
2016-01-12  2:24                       ` Hao, Xudong
2016-01-07 13:28   ` [Qemu-devel] " Cao jin
2016-01-07 15:01     ` Stefano Stabellini
2016-01-07 16:21       ` Gerd Hoffmann
2016-01-07 17:42         ` Stefano Stabellini

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=20151229130853-mutt-send-email-mst@redhat.com \
    --to=mst@redhat.com \
    --cc=caoj.fnst@cn.fujitsu.com \
    --cc=qemu-devel@nongnu.org \
    --cc=stefano.stabellini@eu.citrix.com \
    /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).