linux中的MSI中断。。其实真相是个悲剧

前边说到想用MSI，因为MSI能够提供多样化的中断信息，也就是说一个中断号，却能传递多种信息。这样美好的特性当然是大家所向往的。

可惜，上周看MSI用法的时候，“最终知道真相的我眼泪掉下来”。。T T，真相就是目前linux的内核不支持这种多重MSI（Mutiple MSI）！也就是说MSI使用起来在效果上和传统的irq没区别，都是只能处理一种中断。

看看代码里怎么说。在msi-HOWTO.txt里提到了，要使用msi机制，需要先调用pci_enable_msi接口。这个接口在pci.h中定义

#define pci_enable_msi(pdev)    pci_enable_msi_block(pdev, 1)

这里看到，实际上它调用了pci_enable_msi_block接口。同样在msi-HOWTO中也提到了这个函数，有如下说明：

int pci_enable_msi_block(struct pci_dev *dev, int count)

This variation on the above call allows a device driver to request multiple
MSIs.  The MSI specification only allows interrupts to be allocated in
powers of two, up to a maximum of 2^5 (32).

这里提到了，这个函数其实是用来申请mutiple MSIs的。当时看到这里的时候，我真心想，我的病有救了啊！于是继续深入，位于msi.c中：

/**
 * pci_enable_msi_block - configure device's MSI capability structure
 * @dev: device to configure
 * @nvec: number of interrupts to configure
 *
 * Allocate IRQs for a device with the MSI capability.
 * This function returns a negative errno if an error occurs.  If it
 * is unable to allocate the number of interrupts requested, it returns
 * the number of interrupts it might be able to allocate.  If it successfully
 * allocates at least the number of interrupts requested, it returns 0 and
 * updates the @dev's irq member to the lowest new interrupt number; the
 * other interrupt numbers allocated to this device are consecutive.
 */
int pci_enable_msi_block(struct pci_dev *dev, unsigned int nvec)
{
    int status, pos, maxvec;
    u16 msgctl;

    pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
    if (!pos)
        return -EINVAL;
    pci_read_config_word(dev, pos + PCI_MSI_FLAGS, &msgctl);
    maxvec = 1 << ((msgctl & PCI_MSI_FLAGS_QMASK) >> 1);
    if (nvec > maxvec)
        return maxvec;

    status = pci_msi_check_device(dev, nvec, PCI_CAP_ID_MSI);
    if (status)
        return status;

    WARN_ON(!!dev->msi_enabled);

    /* Check whether driver already requested MSI-X irqs */
    if (dev->msix_enabled) {
        dev_info(&dev->dev, "can't enable MSI "
             "(MSI-X already enabled)\n");
        return -EINVAL;
    }

    status = msi_capability_init(dev, nvec);
    return status;
}

互略前边设备检查的部分，重点就在于 status = msi_capability_init(dev, nvec);这一句。

这个函数做了什么呢。。

static int msi_capability_init(struct pci_dev *dev, int nvec)
{
    struct msi_desc *entry;
    int pos, ret;
    u16 control;
    unsigned mask;

    pos = pci_find_capability(dev, PCI_CAP_ID_MSI);
    msi_set_enable(dev, pos, 0);    /* Disable MSI during set up */

    pci_read_config_word(dev, msi_control_reg(pos), &control);
    /* MSI Entry Initialization */
    entry = alloc_msi_entry(dev);
    if (!entry)
        return -ENOMEM;

    entry->msi_attrib.is_msix    = 0;
    entry->msi_attrib.is_64        = is_64bit_address(control);
    entry->msi_attrib.entry_nr    = 0;
    entry->msi_attrib.maskbit    = is_mask_bit_support(control);
    entry->msi_attrib.default_irq    = dev->irq;    /* Save IOAPIC IRQ */
    entry->msi_attrib.pos        = pos;

    entry->mask_pos = msi_mask_reg(pos, entry->msi_attrib.is_64);
    /* All MSIs are unmasked by default, Mask them all */
    if (entry->msi_attrib.maskbit)
        pci_read_config_dword(dev, entry->mask_pos, &entry->masked);
    mask = msi_capable_mask(control);
    msi_mask_irq(entry, mask, mask);

    list_add_tail(&entry->list, &dev->msi_list);

    /* Configure MSI capability structure */
    ret = arch_setup_msi_irqs(dev, nvec, PCI_CAP_ID_MSI);
    if (ret) {
        msi_mask_irq(entry, mask, ~mask);
        free_msi_irqs(dev);
        return ret;
    }

    /* Set MSI enabled bits     */
    pci_intx_for_msi(dev, 0);
    msi_set_enable(dev, pos, 1);
    dev->msi_enabled = 1;

    dev->irq = entry->irq;
    return 0;
}

可以看到，主要还是先分配和初始化了一些数据结构，并添加到内核维护的链表上。

狗血的高潮在于 ret = arch_setup_msi_irqs(dev, nvec, PCI_CAP_ID_MSI);

字面来看这个函数应该是根据特定architecture建立起msi中断，应该已经接近这番探索的终点了。而事实上也确实是这样。。

代码如下：

#ifndef arch_setup_msi_irqs
int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
{
    struct msi_desc *entry;
    int ret;

    /*
     * If an architecture wants to support multiple MSI, it needs to
     * override arch_setup_msi_irqs()
     */
    if (type == PCI_CAP_ID_MSI && nvec > 1)
        return 1;

    list_for_each_entry(entry, &dev->msi_list, list) {
        ret = arch_setup_msi_irq(dev, entry);
        if (ret < 0)
            return ret;
        if (ret > 0)
            return -ENOSPC;
    }

    return 0;
}
#endif

看到了吗。。。。当nvec>1时候直接就return 1.。。。解释就是如果想支持mutiple MSI，请客观您自行重载该函数。。。