From: Russ Weight <russell.h.weight@intel.com>
To: Xu Yilun <yilun.xu@intel.com>
Cc: Tom Rix <trix@redhat.com>,
mdf@kernel.org, linux-fpga@vger.kernel.org, lgoncalv@redhat.com,
hao.wu@intel.com, matthew.gerlach@intel.com,
richard.gong@intel.com
Subject: Re: [PATCH v4 1/3] fpga: mgr: Use standard dev_release for class driver
Date: Mon, 7 Jun 2021 16:50:58 -0700 [thread overview]
Message-ID: <6ba648a9-0722-856d-0a4f-a2b8eba68355@intel.com> (raw)
In-Reply-To: <20210605154253.GA1973240@yilunxu-OptiPlex-7050>
On 6/5/21 8:42 AM, Xu Yilun wrote:
> On Thu, Jun 03, 2021 at 03:45:17PM -0700, Russ Weight wrote:
>>
>> On 6/1/21 5:38 PM, Russ Weight wrote:
>>> On 6/1/21 4:24 PM, Tom Rix wrote:
>>>> On 5/28/21 2:57 PM, Russ Weight wrote:
>>>>> The FPGA manager class driver data structure is being treated as a
>>>>> managed resource instead of using standard dev_release call-back
>>>>> to release the class data structure. This change removes the
>>>>> managed resource code for the freeing of the class data structure
>>>>> and combines the create() and register() functions into a single
>>>>> register() function.
>>>>>
>>>>> The devm_fpga_mgr_register() function is retained.
>>>>>
>>>>> Signed-off-by: Russ Weight <russell.h.weight@intel.com>
>>>>> ---
>>>>> v4:
>>>>> - Added the compat_id parameter to fpga_mgr_register() and
>>>> On a refactor, it is best to do the refactoring as mechanically as possible.
>>>>
>>>> While compat_id may be useful, it should be done in a separate patch.
>>> These would have to be included with the changes merge the create and register
>>> functions into a single function.
>>>
>>>> Unless it is really needed for your next security update patch consider resubmitting later
>>>>
>>>>> devm_fpga_mgr_register() to ensure that the compat_id is set before
>>>>> the device_register() call.
>>>>> v3:
>>>>> - Cleaned up comment header for fpga_mgr_register()
>>>>> - Fix error return on ida_simple_get() failure
>>>>> v2:
>>>>> - Restored devm_fpga_mgr_register() functionality, adapted for the combined
>>>>> create/register functionality.
>>>>> - All previous callers of devm_fpga_mgr_register() will continue to call
>>>>> devm_fpga_mgr_register().
>>>>> - replaced unnecessary ternary operators in return statements with standard
>>>>> if conditions.
>>>>> ---
>>>>> drivers/fpga/altera-cvp.c | 12 +--
>>>>> drivers/fpga/altera-pr-ip-core.c | 9 +-
>>>>> drivers/fpga/altera-ps-spi.c | 10 +-
>>>>> drivers/fpga/dfl-fme-mgr.c | 12 +--
>>>>> drivers/fpga/fpga-mgr.c | 168 +++++++++----------------------
>>>>> drivers/fpga/ice40-spi.c | 10 +-
>>>>> drivers/fpga/machxo2-spi.c | 10 +-
>>>>> drivers/fpga/socfpga-a10.c | 16 ++-
>>>>> drivers/fpga/socfpga.c | 10 +-
>>>>> drivers/fpga/stratix10-soc.c | 15 +--
>>>>> drivers/fpga/ts73xx-fpga.c | 10 +-
>>>>> drivers/fpga/xilinx-spi.c | 12 +--
>>>>> drivers/fpga/zynq-fpga.c | 16 ++-
>>>>> drivers/fpga/zynqmp-fpga.c | 10 +-
>>>>> include/linux/fpga/fpga-mgr.h | 19 ++--
>>>>> 15 files changed, 120 insertions(+), 219 deletions(-)
>>>> ./Documentation/driver-api/fpga/fpga-mgr.rst
>>>>
>>>> still refers to fpga_mgr_create.
>>> Thanks for catching that.
>>>
>>>>> diff --git a/drivers/fpga/altera-cvp.c b/drivers/fpga/altera-cvp.c
>>>>> index 4e0edb60bfba..25eaa97ae7aa 100644
>>>>> --- a/drivers/fpga/altera-cvp.c
>>>>> +++ b/drivers/fpga/altera-cvp.c
>>>>> @@ -652,19 +652,15 @@ static int altera_cvp_probe(struct pci_dev *pdev,
>>>>> snprintf(conf->mgr_name, sizeof(conf->mgr_name), "%s @%s",
>>>>> ALTERA_CVP_MGR_NAME, pci_name(pdev));
>>>>> - mgr = devm_fpga_mgr_create(&pdev->dev, conf->mgr_name,
>>>>> - &altera_cvp_ops, conf);
>>>>> - if (!mgr) {
>>>>> - ret = -ENOMEM;
>>>>> + mgr = fpga_mgr_register(&pdev->dev, conf->mgr_name, NULL,
>>>>> + &altera_cvp_ops, conf);
>>>>> + if (IS_ERR(mgr)) {
>>>>> + ret = PTR_ERR(mgr);
>>>>> goto err_unmap;
>>>>> }
>>>>> pci_set_drvdata(pdev, mgr);
>>>>> - ret = fpga_mgr_register(mgr);
>>>>> - if (ret)
>>>>> - goto err_unmap;
>>>>> -
>>>>> return 0;
>>>>> err_unmap:
>>>>> diff --git a/drivers/fpga/altera-pr-ip-core.c b/drivers/fpga/altera-pr-ip-core.c
>>>>> index 5b130c4d9882..bfd4de1e7ac5 100644
>>>>> --- a/drivers/fpga/altera-pr-ip-core.c
>>>>> +++ b/drivers/fpga/altera-pr-ip-core.c
>>>>> @@ -191,11 +191,12 @@ int alt_pr_register(struct device *dev, void __iomem *reg_base)
>>>>> (val & ALT_PR_CSR_STATUS_MSK) >> ALT_PR_CSR_STATUS_SFT,
>>>>> (int)(val & ALT_PR_CSR_PR_START));
>>>>> - mgr = devm_fpga_mgr_create(dev, dev_name(dev), &alt_pr_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(dev, dev_name(dev), NULL,
>>>>> + &alt_pr_ops, priv);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> EXPORT_SYMBOL_GPL(alt_pr_register);
>>>>> diff --git a/drivers/fpga/altera-ps-spi.c b/drivers/fpga/altera-ps-spi.c
>>>>> index 23bfd4d1ad0f..e6fc7208213c 100644
>>>>> --- a/drivers/fpga/altera-ps-spi.c
>>>>> +++ b/drivers/fpga/altera-ps-spi.c
>>>>> @@ -302,12 +302,12 @@ static int altera_ps_probe(struct spi_device *spi)
>>>>> snprintf(conf->mgr_name, sizeof(conf->mgr_name), "%s %s",
>>>>> dev_driver_string(&spi->dev), dev_name(&spi->dev));
>>>>> - mgr = devm_fpga_mgr_create(&spi->dev, conf->mgr_name,
>>>>> - &altera_ps_ops, conf);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(&spi->dev, conf->mgr_name, NULL,
>>>>> + &altera_ps_ops, conf);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(&spi->dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> static const struct spi_device_id altera_ps_spi_ids[] = {
>>>>> diff --git a/drivers/fpga/dfl-fme-mgr.c b/drivers/fpga/dfl-fme-mgr.c
>>>>> index d5861d13b306..c9b98afa7d50 100644
>>>>> --- a/drivers/fpga/dfl-fme-mgr.c
>>>>> +++ b/drivers/fpga/dfl-fme-mgr.c
>>>>> @@ -308,14 +308,12 @@ static int fme_mgr_probe(struct platform_device *pdev)
>>>>> fme_mgr_get_compat_id(priv->ioaddr, compat_id);
>>>>> - mgr = devm_fpga_mgr_create(dev, "DFL FME FPGA Manager",
>>>>> - &fme_mgr_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> -
>>>>> - mgr->compat_id = compat_id;
>>>>> + mgr = devm_fpga_mgr_register(dev, "DFL FME FPGA Manager", compat_id,
>>>>> + &fme_mgr_ops, priv);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> static struct platform_driver fme_mgr_driver = {
>>>>> diff --git a/drivers/fpga/fpga-mgr.c b/drivers/fpga/fpga-mgr.c
>>>>> index b85bc47c91a9..d6d64ce34417 100644
>>>>> --- a/drivers/fpga/fpga-mgr.c
>>>>> +++ b/drivers/fpga/fpga-mgr.c
>>>>> @@ -550,20 +550,22 @@ void fpga_mgr_unlock(struct fpga_manager *mgr)
>>>>> EXPORT_SYMBOL_GPL(fpga_mgr_unlock);
>>>>> /**
>>>>> - * fpga_mgr_create - create and initialize a FPGA manager struct
>>>>> + * fpga_mgr_register - create and register a FPGA Manager device
>>>>> * @dev: fpga manager device from pdev
>>>>> * @name: fpga manager name
>>>>> + * @compat_id: FPGA manager id for compatibility check.
>>>> for consistency, use lowercase fpga
>>>>
>>>> similar issue below.
>>> OK
>>>
>>>>> * @mops: pointer to structure of fpga manager ops
>>>>> * @priv: fpga manager private data
>>>>> *
>>>>> - * The caller of this function is responsible for freeing the struct with
>>>>> - * fpga_mgr_free(). Using devm_fpga_mgr_create() instead is recommended.
>>>>> + * The caller of this function is responsible for calling fpga_mgr_unregister().
>>>>> + * Using devm_fpga_mgr_register instead is recommended.
>>>>> *
>>>>> - * Return: pointer to struct fpga_manager or NULL
>>>>> + * Return: struct fpga_manager pointer on success, ERR_PTR() on error.
>>>>> */
>>>>> -struct fpga_manager *fpga_mgr_create(struct device *dev, const char *name,
>>>>> - const struct fpga_manager_ops *mops,
>>>>> - void *priv)
>>>>> +struct fpga_manager *
>>>>> +fpga_mgr_register(struct device *dev, const char *name,
>>>>> + struct fpga_compat_id *compat_id,
>>>>> + const struct fpga_manager_ops *mops, void *priv)
>>>>> {
>>>>> struct fpga_manager *mgr;
>>>>> int id, ret;
>>>>> @@ -572,29 +574,31 @@ struct fpga_manager *fpga_mgr_create(struct device *dev, const char *name,
>>>>> !mops->write_init || (!mops->write && !mops->write_sg) ||
>>>>> (mops->write && mops->write_sg)) {
>>>>> dev_err(dev, "Attempt to register without fpga_manager_ops\n");
>>>>> - return NULL;
>>>>> + return ERR_PTR(-EINVAL);
>>>>> }
>>>>> if (!name || !strlen(name)) {
>>>>> dev_err(dev, "Attempt to register with no name!\n");
>>>>> - return NULL;
>>>>> + return ERR_PTR(-EINVAL);
>>>>> }
>>>>> mgr = kzalloc(sizeof(*mgr), GFP_KERNEL);
>>>>> if (!mgr)
>>>>> - return NULL;
>>>>> + return ERR_PTR(-ENOMEM);
>>>>> id = ida_simple_get(&fpga_mgr_ida, 0, 0, GFP_KERNEL);
>>>>> - if (id < 0)
>>>>> + if (id < 0) {
>>>>> + ret = id;
>>>> ida_* returns a normal error, so needs a wrapping in
>>>>
>>>> ret = ERR_PTR(id)
>>> Thanks - I'll fix that
>> Actually, if you follow the "goto error_kfree", you will see that ret is
>> wrapped in ERR_PTR() in the return statement, so id does not need to be
>> wrapped here.
>>
>>>>> goto error_kfree;
>>>>> + }
>>>>> mutex_init(&mgr->ref_mutex);
>>>>> mgr->name = name;
>>>>> + mgr->compat_id = compat_id;
>>>>> mgr->mops = mops;
>>>>> mgr->priv = priv;
>>>>> - device_initialize(&mgr->dev);
>>>>> mgr->dev.class = fpga_mgr_class;
>>>>> mgr->dev.groups = mops->groups;
>>>>> mgr->dev.parent = dev;
>>>>> @@ -605,84 +609,6 @@ struct fpga_manager *fpga_mgr_create(struct device *dev, const char *name,
>>>>> if (ret)
>>>>> goto error_device;
>>>>> - return mgr;
>>>>> -
>>>>> -error_device:
>>>>> - ida_simple_remove(&fpga_mgr_ida, id);
>>>>> -error_kfree:
>>>>> - kfree(mgr);
>>>>> -
>>>>> - return NULL;
>>>>> -}
>>>>> -EXPORT_SYMBOL_GPL(fpga_mgr_create);
>>>>> -
>>>>> -/**
>>>>> - * fpga_mgr_free - free a FPGA manager created with fpga_mgr_create()
>>>>> - * @mgr: fpga manager struct
>>>>> - */
>>>>> -void fpga_mgr_free(struct fpga_manager *mgr)
>>>>> -{
>>>>> - ida_simple_remove(&fpga_mgr_ida, mgr->dev.id);
>>>>> - kfree(mgr);
>>>>> -}
>>>>> -EXPORT_SYMBOL_GPL(fpga_mgr_free);
>>>>> -
>>>>> -static void devm_fpga_mgr_release(struct device *dev, void *res)
>>>>> -{
>>>>> - struct fpga_mgr_devres *dr = res;
>>>>> -
>>>>> - fpga_mgr_free(dr->mgr);
>>>>> -}
>>>>> -
>>>>> -/**
>>>>> - * devm_fpga_mgr_create - create and initialize a managed FPGA manager struct
>>>>> - * @dev: fpga manager device from pdev
>>>>> - * @name: fpga manager name
>>>>> - * @mops: pointer to structure of fpga manager ops
>>>>> - * @priv: fpga manager private data
>>>>> - *
>>>>> - * This function is intended for use in a FPGA manager driver's probe function.
>>>>> - * After the manager driver creates the manager struct with
>>>>> - * devm_fpga_mgr_create(), it should register it with fpga_mgr_register(). The
>>>>> - * manager driver's remove function should call fpga_mgr_unregister(). The
>>>>> - * manager struct allocated with this function will be freed automatically on
>>>>> - * driver detach. This includes the case of a probe function returning error
>>>>> - * before calling fpga_mgr_register(), the struct will still get cleaned up.
>>>>> - *
>>>>> - * Return: pointer to struct fpga_manager or NULL
>>>>> - */
>>>>> -struct fpga_manager *devm_fpga_mgr_create(struct device *dev, const char *name,
>>>>> - const struct fpga_manager_ops *mops,
>>>>> - void *priv)
>>>>> -{
>>>>> - struct fpga_mgr_devres *dr;
>>>>> -
>>>>> - dr = devres_alloc(devm_fpga_mgr_release, sizeof(*dr), GFP_KERNEL);
>>>>> - if (!dr)
>>>>> - return NULL;
>>>>> -
>>>>> - dr->mgr = fpga_mgr_create(dev, name, mops, priv);
>>>>> - if (!dr->mgr) {
>>>>> - devres_free(dr);
>>>>> - return NULL;
>>>>> - }
>>>>> -
>>>>> - devres_add(dev, dr);
>>>>> -
>>>>> - return dr->mgr;
>>>>> -}
>>>>> -EXPORT_SYMBOL_GPL(devm_fpga_mgr_create);
>>>>> -
>>>>> -/**
>>>>> - * fpga_mgr_register - register a FPGA manager
>>>>> - * @mgr: fpga manager struct
>>>>> - *
>>>>> - * Return: 0 on success, negative error code otherwise.
>>>>> - */
>>>>> -int fpga_mgr_register(struct fpga_manager *mgr)
>>>>> -{
>>>>> - int ret;
>>>>> -
>>>>> /*
>>>>> * Initialize framework state by requesting low level driver read state
>>>>> * from device. FPGA may be in reset mode or may have been programmed
>>>>> @@ -690,18 +616,20 @@ int fpga_mgr_register(struct fpga_manager *mgr)
>>>>> */
>>>>> mgr->state = mgr->mops->state(mgr);
>>>>> - ret = device_add(&mgr->dev);
>>>>> - if (ret)
>>>>> - goto error_device;
>>>>> -
>>>>> - dev_info(&mgr->dev, "%s registered\n", mgr->name);
>>>> can we keep this dev_info() line and move it below device_register() ?
>>> In Greg's response to my driver, he said he said that a well-behaved driver does
>>> not need dev_info() calls...
>>>>> + ret = device_register(&mgr->dev);
>>>>> + if (ret) {
>>>>> + put_device(&mgr->dev);
>>>> pr_err() here ?
>>> OK - I can add a message
>>>
>>>>> + return ERR_PTR(ret);
>>>>> + }
>>>>> - return 0;
>>>>> + return mgr;
>>>>> error_device:
>>>>> - ida_simple_remove(&fpga_mgr_ida, mgr->dev.id);
>>>>> + ida_simple_remove(&fpga_mgr_ida, id);
>>>>> +error_kfree:
>>>>> + kfree(mgr);
>>>>> - return ret;
>>>>> + return ERR_PTR(ret);
>>>>> }
>>>>> EXPORT_SYMBOL_GPL(fpga_mgr_register);
>>>>> @@ -726,14 +654,6 @@ void fpga_mgr_unregister(struct fpga_manager *mgr)
>>>>> }
>>>>> EXPORT_SYMBOL_GPL(fpga_mgr_unregister);
>>>>> -static int fpga_mgr_devres_match(struct device *dev, void *res,
>>>>> - void *match_data)
>>>>> -{
>>>>> - struct fpga_mgr_devres *dr = res;
>>>>> -
>>>>> - return match_data == dr->mgr;
>>>>> -}
>>>>> -
>>>>> static void devm_fpga_mgr_unregister(struct device *dev, void *res)
>>>>> {
>>>>> struct fpga_mgr_devres *dr = res;
>>>>> @@ -743,44 +663,46 @@ static void devm_fpga_mgr_unregister(struct device *dev, void *res)
>>>>> /**
>>>>> * devm_fpga_mgr_register - resource managed variant of fpga_mgr_register()
>>>>> - * @dev: managing device for this FPGA manager
>>>>> - * @mgr: fpga manager struct
>>>>> + * @dev: fpga manager device from pdev
>>>>> + * @name: fpga manager name
>>>>> + * @compat_id: FPGA manager id for compatibility check.
>>>>> + * @mops: pointer to structure of fpga manager ops
>>>>> + * @priv: fpga manager private data
>>>>> *
>>>>> * This is the devres variant of fpga_mgr_register() for which the unregister
>>>>> * function will be called automatically when the managing device is detached.
>>>>> */
>>>>> -int devm_fpga_mgr_register(struct device *dev, struct fpga_manager *mgr)
>>>>> +struct fpga_manager *
>>>>> +devm_fpga_mgr_register(struct device *dev, const char *name,
>>>>> + struct fpga_compat_id *compat_id,
>>>>> + const struct fpga_manager_ops *mops, void *priv)
>>>>> {
>>>>> struct fpga_mgr_devres *dr;
>>>>> - int ret;
>>>>> -
>>>>> - /*
>>>>> - * Make sure that the struct fpga_manager * that is passed in is
>>>>> - * managed itself.
>>>>> - */
>>>>> - if (WARN_ON(!devres_find(dev, devm_fpga_mgr_release,
>>>>> - fpga_mgr_devres_match, mgr)))
>>>>> - return -EINVAL;
>>>>> + struct fpga_manager *mgr;
>>>>> dr = devres_alloc(devm_fpga_mgr_unregister, sizeof(*dr), GFP_KERNEL);
>>>>> if (!dr)
>>>>> - return -ENOMEM;
>>>>> + return ERR_PTR(-ENOMEM);
>>>>> - ret = fpga_mgr_register(mgr);
>>>>> - if (ret) {
>>>>> + mgr = fpga_mgr_register(dev, name, compat_id, mops, priv);
>>>>> + if (IS_ERR(mgr)) {
>>>>> devres_free(dr);
>>>>> - return ret;
>>>>> + return mgr;
>>>>> }
>>>>> dr->mgr = mgr;
>>>>> devres_add(dev, dr);
>>>>> - return 0;
>>>>> + return mgr;
>>>>> }
>>>>> EXPORT_SYMBOL_GPL(devm_fpga_mgr_register);
>>>>> static void fpga_mgr_dev_release(struct device *dev)
>>>>> {
>>>>> + struct fpga_manager *mgr = to_fpga_manager(dev);
>>>>> +
>>>>> + ida_simple_remove(&fpga_mgr_ida, mgr->dev.id);
>>>>> + kfree(mgr);
>>>>> }
>>>>> static int __init fpga_mgr_class_init(void)
>>>>> diff --git a/drivers/fpga/ice40-spi.c b/drivers/fpga/ice40-spi.c
>>>>> index 69dec5af23c3..b1c7c0713e11 100644
>>>>> --- a/drivers/fpga/ice40-spi.c
>>>>> +++ b/drivers/fpga/ice40-spi.c
>>>>> @@ -178,12 +178,12 @@ static int ice40_fpga_probe(struct spi_device *spi)
>>>>> return ret;
>>>>> }
>>>>> - mgr = devm_fpga_mgr_create(dev, "Lattice iCE40 FPGA Manager",
>>>>> - &ice40_fpga_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(dev, "Lattice iCE40 FPGA Manager",
>>>>> + NULL, &ice40_fpga_ops, priv);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> static const struct of_device_id ice40_fpga_of_match[] = {
>>>>> diff --git a/drivers/fpga/machxo2-spi.c b/drivers/fpga/machxo2-spi.c
>>>>> index 114a64d2b7a4..75cc6efaa1b6 100644
>>>>> --- a/drivers/fpga/machxo2-spi.c
>>>>> +++ b/drivers/fpga/machxo2-spi.c
>>>>> @@ -366,12 +366,12 @@ static int machxo2_spi_probe(struct spi_device *spi)
>>>>> return -EINVAL;
>>>>> }
>>>>> - mgr = devm_fpga_mgr_create(dev, "Lattice MachXO2 SPI FPGA Manager",
>>>>> - &machxo2_ops, spi);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(dev, "Lattice MachXO2 SPI FPGA Manager",
>>>>> + NULL, &machxo2_ops, spi);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> static const struct of_device_id of_match[] = {
>>>>> diff --git a/drivers/fpga/socfpga-a10.c b/drivers/fpga/socfpga-a10.c
>>>>> index 573d88bdf730..80b8891fb302 100644
>>>>> --- a/drivers/fpga/socfpga-a10.c
>>>>> +++ b/drivers/fpga/socfpga-a10.c
>>>>> @@ -508,19 +508,15 @@ static int socfpga_a10_fpga_probe(struct platform_device *pdev)
>>>>> return -EBUSY;
>>>>> }
>>>>> - mgr = devm_fpga_mgr_create(dev, "SoCFPGA Arria10 FPGA Manager",
>>>>> - &socfpga_a10_fpga_mgr_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> -
>>>>> - platform_set_drvdata(pdev, mgr);
>>>>> -
>>>>> - ret = fpga_mgr_register(mgr);
>>>>> - if (ret) {
>>>>> + mgr = fpga_mgr_register(dev, "SoCFPGA Arria10 FPGA Manager",
>>>>> + NULL, &socfpga_a10_fpga_mgr_ops, priv);
>>>>> + if (IS_ERR(mgr)) {
>>>>> clk_disable_unprepare(priv->clk);
>>>>> - return ret;
>>>>> + return PTR_ERR(mgr);
>>>>> }
>>>>> + platform_set_drvdata(pdev, mgr);
>>>>> +
>>>>> return 0;
>>>>> }
>>>>> diff --git a/drivers/fpga/socfpga.c b/drivers/fpga/socfpga.c
>>>>> index 1f467173fc1f..06edfab8a496 100644
>>>>> --- a/drivers/fpga/socfpga.c
>>>>> +++ b/drivers/fpga/socfpga.c
>>>>> @@ -571,12 +571,12 @@ static int socfpga_fpga_probe(struct platform_device *pdev)
>>>>> if (ret)
>>>>> return ret;
>>>>> - mgr = devm_fpga_mgr_create(dev, "Altera SOCFPGA FPGA Manager",
>>>>> - &socfpga_fpga_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(dev, "Altera SOCFPGA FPGA Manager",
>>>>> + NULL, &socfpga_fpga_ops, priv);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> #ifdef CONFIG_OF
>>>>> diff --git a/drivers/fpga/stratix10-soc.c b/drivers/fpga/stratix10-soc.c
>>>>> index 657a70c5fc99..434094c3ac69 100644
>>>>> --- a/drivers/fpga/stratix10-soc.c
>>>>> +++ b/drivers/fpga/stratix10-soc.c
>>>>> @@ -425,18 +425,11 @@ static int s10_probe(struct platform_device *pdev)
>>>>> init_completion(&priv->status_return_completion);
>>>>> - mgr = fpga_mgr_create(dev, "Stratix10 SOC FPGA Manager",
>>>>> - &s10_ops, priv);
>>>>> - if (!mgr) {
>>>>> - dev_err(dev, "unable to create FPGA manager\n");
>>>>> - ret = -ENOMEM;
>>>>> - goto probe_err;
>>>>> - }
>>>>> -
>>>>> - ret = fpga_mgr_register(mgr);
>>>>> - if (ret) {
>>>>> + mgr = fpga_mgr_register(dev, "Stratix10 SOC FPGA Manager",
>>>>> + NULL, &s10_ops, priv);
>>>>> + if (IS_ERR(mgr)) {
>>>>> dev_err(dev, "unable to register FPGA manager\n");
>>>>> - fpga_mgr_free(mgr);
>>>>> + ret = PTR_ERR(mgr);
>>>>> goto probe_err;
>>>>> }
>>>>> diff --git a/drivers/fpga/ts73xx-fpga.c b/drivers/fpga/ts73xx-fpga.c
>>>>> index 101f016c6ed8..5efbec2b9f9f 100644
>>>>> --- a/drivers/fpga/ts73xx-fpga.c
>>>>> +++ b/drivers/fpga/ts73xx-fpga.c
>>>>> @@ -122,12 +122,12 @@ static int ts73xx_fpga_probe(struct platform_device *pdev)
>>>>> if (IS_ERR(priv->io_base))
>>>>> return PTR_ERR(priv->io_base);
>>>>> - mgr = devm_fpga_mgr_create(kdev, "TS-73xx FPGA Manager",
>>>>> - &ts73xx_fpga_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(kdev, "TS-73xx FPGA Manager",
>>>> This is missing compat_id param.
>>> Thanks - I thought I'll do another pass and make sure I catch them all.
>>>
>>>> Similar for others.
>>>>
>>>>> + &ts73xx_fpga_ops, priv);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(kdev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> static struct platform_driver ts73xx_fpga_driver = {
>>>>> diff --git a/drivers/fpga/xilinx-spi.c b/drivers/fpga/xilinx-spi.c
>>>>> index fee4d0abf6bf..6f1894929687 100644
>>>>> --- a/drivers/fpga/xilinx-spi.c
>>>>> +++ b/drivers/fpga/xilinx-spi.c
>>>>> @@ -247,13 +247,13 @@ static int xilinx_spi_probe(struct spi_device *spi)
>>>>> return dev_err_probe(&spi->dev, PTR_ERR(conf->done),
>>>>> "Failed to get DONE gpio\n");
>>>>> - mgr = devm_fpga_mgr_create(&spi->dev,
>>>>> - "Xilinx Slave Serial FPGA Manager",
>>>>> - &xilinx_spi_ops, conf);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(&spi->dev,
>>>>> + "Xilinx Slave Serial FPGA Manager",
>>>>> + NULL, &xilinx_spi_ops, conf);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(&spi->dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> static const struct of_device_id xlnx_spi_of_match[] = {
>>>>> diff --git a/drivers/fpga/zynq-fpga.c b/drivers/fpga/zynq-fpga.c
>>>>> index 07fa8d9ec675..74bbb9710f4e 100644
>>>>> --- a/drivers/fpga/zynq-fpga.c
>>>>> +++ b/drivers/fpga/zynq-fpga.c
>>>>> @@ -609,20 +609,16 @@ static int zynq_fpga_probe(struct platform_device *pdev)
>>>>> clk_disable(priv->clk);
>>>>> - mgr = devm_fpga_mgr_create(dev, "Xilinx Zynq FPGA Manager",
>>>>> - &zynq_fpga_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> -
>>>>> - platform_set_drvdata(pdev, mgr);
>>>>> -
>>>>> - err = fpga_mgr_register(mgr);
>>>>> - if (err) {
>>>>> + mgr = fpga_mgr_register(dev, "Xilinx Zynq FPGA Manager",
>>>> This is missing compat_id param as well.
>>>>> + &zynq_fpga_ops, priv);
>>>>> + if (IS_ERR(mgr)) {
>>>>> dev_err(dev, "unable to register FPGA manager\n");
>>>>> clk_unprepare(priv->clk);
>>>>> - return err;
>>>>> + return PTR_ERR(mgr);
>>>>> }
>>>>> + platform_set_drvdata(pdev, mgr);
>>>>> +
>>>>> return 0;
>>>>> }
>>>>> diff --git a/drivers/fpga/zynqmp-fpga.c b/drivers/fpga/zynqmp-fpga.c
>>>>> index 125743c9797f..1a6d520985be 100644
>>>>> --- a/drivers/fpga/zynqmp-fpga.c
>>>>> +++ b/drivers/fpga/zynqmp-fpga.c
>>>>> @@ -102,12 +102,12 @@ static int zynqmp_fpga_probe(struct platform_device *pdev)
>>>>> priv->dev = dev;
>>>>> - mgr = devm_fpga_mgr_create(dev, "Xilinx ZynqMP FPGA Manager",
>>>>> - &zynqmp_fpga_ops, priv);
>>>>> - if (!mgr)
>>>>> - return -ENOMEM;
>>>>> + mgr = devm_fpga_mgr_register(dev, "Xilinx ZynqMP FPGA Manager",
>>>>> + &zynqmp_fpga_ops, priv);
>>>>> + if (IS_ERR(mgr))
>>>>> + return PTR_ERR(mgr);
>>>>> - return devm_fpga_mgr_register(dev, mgr);
>>>>> + return 0;
>>>>> }
>>>>> static const struct of_device_id zynqmp_fpga_of_match[] = {
>>>>> diff --git a/include/linux/fpga/fpga-mgr.h b/include/linux/fpga/fpga-mgr.h
>>>>> index 2bc3030a69e5..3e1859cc0640 100644
>>>>> --- a/include/linux/fpga/fpga-mgr.h
>>>>> +++ b/include/linux/fpga/fpga-mgr.h
>>>>> @@ -191,17 +191,16 @@ struct fpga_manager *fpga_mgr_get(struct device *dev);
>>>>> void fpga_mgr_put(struct fpga_manager *mgr);
>>>>> -struct fpga_manager *fpga_mgr_create(struct device *dev, const char *name,
>>>>> - const struct fpga_manager_ops *mops,
>>>>> - void *priv);
>>>> My biggest concern is that removing this will break all oot fpga drivers.
>>>>
>>>> The other patches have similar api changes
>>>>
>>>> Having a stable api is very important.
>>> That is a good point. While Greg did call out the separation of the
>>> device_initialize() and device_add() calls, his biggest concern was with
>>> the empty dev_release() functions. Maybe it would be best to first submit
>>> a patch that only addresses the dev_release() functions? Maybe it would be
>>> OK to leave the create and register functions as separate functions in order
>>> to maintain a stable API...
>> On further investigation, I don't think we have much choice in changing
>> the API. We are almost coming full circle here. In 2018, there was only
>> an fpga_mgr_register() function. It was returning an errno value or zero.
>> It was also setting the drvdata of the parent driver with a pointer to
>> the resulting fpga_manager structure.
>>
>> It looks like the reason the interface was split into two (create and
>> register) was so that the fpga_manager pointer could be returned (by the
>> create function) without writing the drvdata of the parent driver:
>>
>> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=7085e2a94f7df5f419e3cfb2fe809ce6564e9629
>>
>> When the interfaces were split, fpga_mgr_free() was also created to free
>> the memory if the register function was not called. The fpga_mgr_free()
>> function was also called by the fpga_mgr_dev_release() function. A later
>> patch was submitted to simplify the memory management for the class driver:
>>
>> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=084181fe8cc7472695b907f0018f4cd00934cb12
>>
>> It is this second commit that empties the fpga_mgr_dev_release() function
>> and instead uses devm_fpga_mgr_create() to manage the memory for the class
>> driver.
>>
>> The fix I am proposing essentially takes us back to the 2018 implementation,
>> but returns the fpga_manager pointer or ERR_PTR() instead of returning zero or
>> an errno.
>>
>> The definition of the API is integral to the memory management issue. I cannot
>> think of an acceptable solution for the upstream that doesn't involve
>> changing the API. It could result in some ifdef's in an out-of-tree driver,
>> but I don't think we can avoid it in this case.
> I agree the stable API is important, and I think it is still possible to fix
> the memory management issue while keeping the APIs.
>
> We should do the actual resource free work in class.dev_release(), and
> fpga_mgr_free should call put_device() to trigger it.
>
> And I see no problem with the current devm APIs.
Consider the case where both devm_fpga_mgr_create() and devm_fpga_mgr_register()
are called, and both succeed. When the driver is unloaded devm_fpga_mgr_release()
is called on behalf of the create and it will call put_device(). Then
devm_fpga_mgr_unregister() will be called on behalf of the register, and it
will call device_unregister(), which also calls put_device(). In this case, I
think there is an extra call to put_device(), right? Or is that not a problem?
The only way I can think to avoid the extra put_device() call for this case
would be to set a flag when-and-if fpga_mgr_register() succeeds and then refuse
to call put_device() in devm_fpga_mgr_release() in order to avoid the duplicate
call to put_device(). But that feels like a hack to me.
Am I missing something? Is there a cleaner way to make this work?
- Russ
>
> Thanks,
> Yilun
>
>>> - Russ
>>>
>>>> My recommendation is to keep the public api and hide the change with an exported but not public function.
>>>>
>>>> Tom
>>>>
>>>>> -void fpga_mgr_free(struct fpga_manager *mgr);
>>>>> -int fpga_mgr_register(struct fpga_manager *mgr);
>>>>> +struct fpga_manager *fpga_mgr_register(struct device *dev, const char *name,
>>>>> + struct fpga_compat_id *compat_id,
>>>>> + const struct fpga_manager_ops *mops,
>>>>> + void *priv);
>>>>> void fpga_mgr_unregister(struct fpga_manager *mgr);
>>>>> -int devm_fpga_mgr_register(struct device *dev, struct fpga_manager *mgr);
>>>>> -
>>>>> -struct fpga_manager *devm_fpga_mgr_create(struct device *dev, const char *name,
>>>>> - const struct fpga_manager_ops *mops,
>>>>> - void *priv);
>>>>> +struct fpga_manager *
>>>>> +devm_fpga_mgr_register(struct device *dev, const char *name,
>>>>> + struct fpga_compat_id *compat_id,
>>>>> + const struct fpga_manager_ops *mops,
>>>>> + void *priv);
>>>>> #endif /*_LINUX_FPGA_MGR_H */
next prev parent reply other threads:[~2021-06-07 23:51 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-28 21:57 [PATCH v4 0/3] fpga: Use standard class dev_release function Russ Weight
2021-05-28 21:57 ` [PATCH v4 1/3] fpga: mgr: Use standard dev_release for class driver Russ Weight
2021-05-31 8:54 ` Xu Yilun
2021-06-03 21:21 ` Russ Weight
2021-06-01 23:24 ` Tom Rix
2021-06-02 0:38 ` Russ Weight
2021-06-03 22:45 ` Russ Weight
2021-06-04 20:56 ` Tom Rix
2021-06-04 21:17 ` Russ Weight
2021-06-05 15:42 ` Xu Yilun
2021-06-07 23:50 ` Russ Weight [this message]
2021-06-08 1:40 ` Xu Yilun
2021-06-08 16:44 ` Russ Weight
2021-05-28 21:57 ` [PATCH v4 2/3] fpga: bridge: " Russ Weight
2021-05-28 21:57 ` [PATCH v4 3/3] fpga: region: " Russ Weight
2021-05-31 15:23 ` Xu Yilun
2021-06-03 22:55 ` Russ Weight
2021-05-31 15:24 ` [PATCH v4 0/3] fpga: Use standard class dev_release function Xu Yilun
2021-06-03 22:49 ` Russ Weight
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=6ba648a9-0722-856d-0a4f-a2b8eba68355@intel.com \
--to=russell.h.weight@intel.com \
--cc=hao.wu@intel.com \
--cc=lgoncalv@redhat.com \
--cc=linux-fpga@vger.kernel.org \
--cc=matthew.gerlach@intel.com \
--cc=mdf@kernel.org \
--cc=richard.gong@intel.com \
--cc=trix@redhat.com \
--cc=yilun.xu@intel.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).