greybus-dev

greybus-dev@lists.linaro.org

9 participants
573 discussions

[PATCH] staging: greybus: fix an error handling bug in gb_svc_hello()

by Dan Carpenter

Cleanup if gb_svc_queue_deferred_request() fails. Fixes: ee2f2074fdb2 ("greybus: svc: reconfig APBridgeA-Switch link to handle required load") Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> --- From static analysis. Not tested. drivers/greybus/svc.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/drivers/greybus/svc.c b/drivers/greybus/svc.c index ce7740ef449b..b19651048081 100644 --- a/drivers/greybus/svc.c +++ b/drivers/greybus/svc.c @@ -866,8 +866,14 @@ static int gb_svc_hello(struct gb_operation *op) gb_svc_debugfs_init(svc); - return gb_svc_queue_deferred_request(op); + ret = gb_svc_queue_deferred_request(op); + if (ret) + goto remove_debugfs; + + return 0; +remove_debugfs: + gb_svc_debugfs_exit(svc); err_unregister_device: gb_svc_watchdog_destroy(svc); device_del(&svc->dev); -- 2.20.1

3 years, 4 months

[PATCH 0/7] Provide and use generic_handle_irq_safe() where appropriate.

by Sebastian Andrzej Siewior

generic_handle_irq() must be used from primary IRQ-handler (chain handler/ interrupt controller entry). It is low level code and the function expects that interrupts are disabled at entry point. This isn't the case for invocations from tasklets, workqueues or the primary interrupt handler on PREEMPT_RT. Once this gets noticed a "local_irq_disable|safe()" is added. To avoid further confusion this series adds generic_handle_irq_safe() which can be used from any context and adds a few user. Sebastian

3 years, 4 months

[PATCH][next] staging: greybus: i2c: Use struct_size() helper in gb_i2c_operation_create()

by Gustavo A. R. Silva

Make use of the struct_size() helper instead of an open-coded version, in order to avoid any potential type mistakes or integer overflows that, in the worst scenario, could lead to heap overflows. Also, address the following sparse warnings: drivers/staging/greybus/i2c.c:111:24: warning: using sizeof on a flexible structure Link: https://github.com/KSPP/linux/issues/174 Signed-off-by: Gustavo A. R. Silva <gustavoars(a)kernel.org> --- drivers/staging/greybus/i2c.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/staging/greybus/i2c.c b/drivers/staging/greybus/i2c.c index de2f6516da09..9dfc6791c20e 100644 --- a/drivers/staging/greybus/i2c.c +++ b/drivers/staging/greybus/i2c.c @@ -108,9 +108,7 @@ gb_i2c_operation_create(struct gb_connection *connection, else data_out_size += (u32)msg->len; - request_size = sizeof(*request); - request_size += msg_count * sizeof(*op); - request_size += data_out_size; + request_size = struct_size(request, ops, msg_count) + data_out_size; /* Response consists only of incoming data */ operation = gb_operation_create(connection, GB_I2C_TYPE_TRANSFER, -- 2.27.0

3 years, 4 months

Re: [greybus-dev] [PATCH] staging: greybus: audio: Check null pointer

by Alex Elder

On 12/27/21 8:02 PM, Jiasheng Jiang wrote: > Sorry the previous email is forgetten to wrap line. > This email is corrected and the content is the same. > > On Mon, Dec 27, 2021 at 11:54:10PM +0800, Alex Elder wrote: >> I think this is a good change, but I would like you to improve >> the description, and fix some different bugs introduced by your >> change. >> >> What you are specifically doing is checking for a null return >> from devm_kcalloc() in gb_generate_enum_strings(), and are >> returning the NULL pointer if that occurs. That means you >> need to update all the callers of gb_generate_enum_strings() >> to also handle a possible null return value. >> >> The fix does a good thing, and your description is correct >> about what you are fixing. But it should supply more >> complete context for the change. > > Thanks for your advice, I will correct my description in next version. > But I still have some question about the devm_kzalloc(). > >> You can't simply return here. If you look a bit above this, >> where the call to allocate a control structure is done, you >> see that a NULL return there jumps to the "error" label, so >> any already allocated and initialized control widgets get >> cleaned up before returning. > > Actually, I have already thought of whether it needs to free after the > devm_kzalloc(). > As we can find in the gbaudio_tplg_create_widget(), the widget_kctls is > allocated by devm_kzalloc(), but isn't released when > gbaudio_tplg_create_wcontrol() fails and goto error. > And I check of the comment of the devm_kmalloc() in `drivers/base/devres.c`, > because devm_kzalloc() returns devm_kmalloc(). > And it says that "Memory allocated with this function is automatically > freed on driver detach." > So there is no need to free the memory manually. > Is that right? You are partially right, but you're missing something. Yes, anything allocated with devm_kzalloc() will be freed automatically when the last reference to the device is dropped. But the two places you're returning are in the middle of a loop (in gbaudio_tplg_create_widget() and gbaudio_tplg_process_kcontrols()). Each is building up a sort of hierarchical data structure, and as you can see, the existing code takes care to de-construct the partially built structure in the event an error occurs before it completes. There is a chance that your simple return would "work", but by reading the surrounding code you should recognize that the proper way to handle the error is to jump to the cleanup at the end. The other alternative would be to fix those functions so they don't do that controlled cleanup on error and simply return early (as you were proposing). But without digging deeper, I would assume the original developers designed it this way very intentionally, because it avoids a problem somewhere. -Alex > And I am sorry again because of the lack of the above explanation in my > commit message. > I will also add to my new commit. > > Thanks, > Jiang >

3 years, 5 months

Re: [PATCH v3] staging: greybus: audio: Check null pointer

by Alex Elder

On 1/4/22 9:06 AM, Jiasheng Jiang wrote: > As the possible alloc failure of devm_kcalloc(), it could return null > pointer. > Therefore, 'strings' should be checked and return NULL if alloc fails to > prevent the dereference of the NULL pointer. > Also, the caller should also deal with the return value of the > gb_generate_enum_strings() and return -ENOMEM if returns NULL. > Moreover, because the memory allocated with devm_kzalloc() will be > freed automatically when the last reference to the device is dropped, > the 'gbe' in gbaudio_tplg_create_enum_kctl() and > gbaudio_tplg_create_enum_ctl() do not need to free manually. > But the 'control' in gbaudio_tplg_create_widget() and > gbaudio_tplg_process_kcontrols() has a specially error handle to > cleanup. > So it should be better to cleanup 'control' when fails. > > Fixes: e65579e335da ("greybus: audio: topology: Enable enumerated control support") > Signed-off-by: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> This looks good to me. Thanks for fixing things. Reviewed-by: Alex Elder <elder(a)linaro.org> > --- > v3: Same code as v2, but remove the redundant message as requested. > v2: Updated to use common error processing at the end of both > gbaudio_tplg_create_widget() and gbaudio_tplg_process_kcontrols(). > --- > drivers/staging/greybus/audio_topology.c | 15 +++++++++++++++ > 1 file changed, 15 insertions(+) > > diff --git a/drivers/staging/greybus/audio_topology.c b/drivers/staging/greybus/audio_topology.c > index 1fc7727ab7be..6bba735d2e5c 100644 > --- a/drivers/staging/greybus/audio_topology.c > +++ b/drivers/staging/greybus/audio_topology.c > @@ -147,6 +147,9 @@ static const char **gb_generate_enum_strings(struct gbaudio_module_info *gb, > > items = le32_to_cpu(gbenum->items); > strings = devm_kcalloc(gb->dev, items, sizeof(char *), GFP_KERNEL); > + if (!strings) > + return NULL; > + > data = gbenum->names; > > for (i = 0; i < items; i++) { > @@ -655,6 +658,8 @@ static int gbaudio_tplg_create_enum_kctl(struct gbaudio_module_info *gb, > /* since count=1, and reg is dummy */ > gbe->items = le32_to_cpu(gb_enum->items); > gbe->texts = gb_generate_enum_strings(gb, gb_enum); > + if (!gbe->texts) > + return -ENOMEM; > > /* debug enum info */ > dev_dbg(gb->dev, "Max:%d, name_length:%d\n", gbe->items, > @@ -862,6 +867,8 @@ static int gbaudio_tplg_create_enum_ctl(struct gbaudio_module_info *gb, > /* since count=1, and reg is dummy */ > gbe->items = le32_to_cpu(gb_enum->items); > gbe->texts = gb_generate_enum_strings(gb, gb_enum); > + if (!gbe->texts) > + return -ENOMEM; > > /* debug enum info */ > dev_dbg(gb->dev, "Max:%d, name_length:%d\n", gbe->items, > @@ -1034,6 +1041,10 @@ static int gbaudio_tplg_create_widget(struct gbaudio_module_info *module, > csize += le16_to_cpu(gbenum->names_length); > control->texts = (const char * const *) > gb_generate_enum_strings(module, gbenum); > + if (!control->texts) { > + ret = -ENOMEM; > + goto error; > + } > control->items = le32_to_cpu(gbenum->items); > } else { > csize = sizeof(struct gb_audio_control); > @@ -1183,6 +1194,10 @@ static int gbaudio_tplg_process_kcontrols(struct gbaudio_module_info *module, > csize += le16_to_cpu(gbenum->names_length); > control->texts = (const char * const *) > gb_generate_enum_strings(module, gbenum); > + if (!control->texts) { > + ret = -ENOMEM; > + goto error; > + } > control->items = le32_to_cpu(gbenum->items); > } else { > csize = sizeof(struct gb_audio_control); >

3 years, 5 months

Re: [PATCH v2] staging: greybus: audio: Check null pointer

by Alex Elder

On 12/28/21 8:28 PM, Jiasheng Jiang wrote: > On Tue, Dec 28, 2021 at 10:50:22PM +0800, Alex Elder wrote: >> But the two places you're returning are in the middle of a loop (in >> gbaudio_tplg_create_widget() and gbaudio_tplg_process_kcontrols()). >> Each is building up a sort of hierarchical data structure, and as >> you can see, the existing code takes care to de-construct the partially >> built structure in the event an error occurs before it completes. >> >> There is a chance that your simple return would "work", but by >> reading the surrounding code you should recognize that the proper >> way to handle the error is to jump to the cleanup at the end. > > Yes, I agree with your opinion and this time I use "goto error" instead > of directly return in the two callers. > The new patch is as follow. > > As the possible alloc failure of devm_kcalloc(), it could return null > pointer. > Therefore, 'strings' should be checked and return NULL if alloc fails to > prevent the dereference of the NULL pointer. > Also, the caller should also deal with the return value of the > gb_generate_enum_strings() and return -ENOMEM if returns NULL. > Moreover, because the memory allocated with devm_kzalloc() will be > freed automatically when the last reference to the device is dropped, > the 'gbe' in gbaudio_tplg_create_enum_kctl() and > gbaudio_tplg_create_enum_ctl() do not need to free manually. > But the 'control' in gbaudio_tplg_create_widget() and > gbaudio_tplg_process_kcontrols() has a specially error handle to > cleanup. > So it should be better to cleanup 'control' when fails. I haven't really looked at this yet, but in any case I would like you to send version 3 of this patch. The reason is that you should not include the message above in the patch itself. To be clear, I really appreciate your response; please *do* respond to review comments. But do so in e-mail only, and then follow up with a new patch, using the same basic subject (with a new version) and (roughly) the same description. So your patch should have a subject line with v3. It should then contain the original description (not indented with ">>" or anything), updated or improved if appropriate to reflect the current state of the patch. Then, below the "---" line you should include a patch history, with a line or two describing each version. For example, it might look something like: --- v3: Same code as v2, but fixed description as requested. v2: Updated to use common error processing at the end of both gbaudio_tplg_create_widget() and gbaudio_tplg_process_kcontrols(). I will review version 3, and if I don't see anything else wrong with it I'll offer a "Signed-off-by" tag. -Alex > Fixes: e65579e335da ("greybus: audio: topology: Enable enumerated control support") > Signed-off-by: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> > --- > Changelog: > > v1 -> v2 > > *Change 1. Add the cleanup process when alloc fails in two callers and > refine the commit message. > --- > drivers/staging/greybus/audio_topology.c | 15 +++++++++++++++ > 1 file changed, 15 insertions(+) > > diff --git a/drivers/staging/greybus/audio_topology.c b/drivers/staging/greybus/audio_topology.c > index 1fc7727ab7be..6bba735d2e5c 100644 > --- a/drivers/staging/greybus/audio_topology.c > +++ b/drivers/staging/greybus/audio_topology.c > @@ -147,6 +147,9 @@ static const char **gb_generate_enum_strings(struct gbaudio_module_info *gb, > > items = le32_to_cpu(gbenum->items); > strings = devm_kcalloc(gb->dev, items, sizeof(char *), GFP_KERNEL); > + if (!strings) > + return NULL; > + > data = gbenum->names; > > for (i = 0; i < items; i++) { > @@ -655,6 +658,8 @@ static int gbaudio_tplg_create_enum_kctl(struct gbaudio_module_info *gb, > /* since count=1, and reg is dummy */ > gbe->items = le32_to_cpu(gb_enum->items); > gbe->texts = gb_generate_enum_strings(gb, gb_enum); > + if (!gbe->texts) > + return -ENOMEM; > > /* debug enum info */ > dev_dbg(gb->dev, "Max:%d, name_length:%d\n", gbe->items, > @@ -862,6 +867,8 @@ static int gbaudio_tplg_create_enum_ctl(struct gbaudio_module_info *gb, > /* since count=1, and reg is dummy */ > gbe->items = le32_to_cpu(gb_enum->items); > gbe->texts = gb_generate_enum_strings(gb, gb_enum); > + if (!gbe->texts) > + return -ENOMEM; > > /* debug enum info */ > dev_dbg(gb->dev, "Max:%d, name_length:%d\n", gbe->items, > @@ -1034,6 +1041,10 @@ static int gbaudio_tplg_create_widget(struct gbaudio_module_info *module, > csize += le16_to_cpu(gbenum->names_length); > control->texts = (const char * const *) > gb_generate_enum_strings(module, gbenum); > + if (!control->texts) { > + ret = -ENOMEM; > + goto error; > + } > control->items = le32_to_cpu(gbenum->items); > } else { > csize = sizeof(struct gb_audio_control); > @@ -1183,6 +1194,10 @@ static int gbaudio_tplg_process_kcontrols(struct gbaudio_module_info *module, > csize += le16_to_cpu(gbenum->names_length); > control->texts = (const char * const *) > gb_generate_enum_strings(module, gbenum); > + if (!control->texts) { > + ret = -ENOMEM; > + goto error; > + } > control->items = le32_to_cpu(gbenum->items); > } else { > csize = sizeof(struct gb_audio_control); >

3 years, 5 months

[PATCH] staging: greybus: auto_manager: use default_groups in kobj_type

by Greg Kroah-Hartman

There are currently 2 ways to create a set of sysfs files for a kobj_type, through the default_attrs field, and the default_groups field. Move the greybus audio code to use default_groups field which has been the preferred way since aa30f47cf666 ("kobject: Add support for default attribute groups to kobj_type") so that we can soon get rid of the obsolete default_attrs field. Cc: Vaibhav Agarwal <vaibhav.sr(a)gmail.com> Cc: Mark Greer <mgreer(a)animalcreek.com> Cc: Johan Hovold <johan(a)kernel.org> Cc: Alex Elder <elder(a)kernel.org> Cc: greybus-dev(a)lists.linaro.org Cc: linux-staging(a)lists.linux.dev Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/staging/greybus/audio_manager_module.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/staging/greybus/audio_manager_module.c b/drivers/staging/greybus/audio_manager_module.c index 525cf8f8394f..0a0f0a394c84 100644 --- a/drivers/staging/greybus/audio_manager_module.c +++ b/drivers/staging/greybus/audio_manager_module.c @@ -142,11 +142,12 @@ static struct attribute *gb_audio_module_default_attrs[] = { &gb_audio_module_op_devices_attribute.attr, NULL, /* need to NULL terminate the list of attributes */ }; +ATTRIBUTE_GROUPS(gb_audio_module_default); static struct kobj_type gb_audio_module_type = { .sysfs_ops = &gb_audio_module_sysfs_ops, .release = gb_audio_module_release, - .default_attrs = gb_audio_module_default_attrs, + .default_groups = gb_audio_module_default_groups, }; static void send_add_uevent(struct gb_audio_manager_module *module) -- 2.34.1

3 years, 5 months

Re: [greybus-dev] [PATCH] staging: greybus: audio: Check null pointer

by Alex Elder

On 12/24/21 3:03 AM, Jiasheng Jiang wrote: > As the possible alloc failure of devm_kcalloc, it could return null > pointer. > To prevent the dereference of the null pointer, it should be checked. I think this is a good change, but I would like you to improve the description, and fix some different bugs introduced by your change. What you are specifically doing is checking for a null return from devm_kcalloc() in gb_generate_enum_strings(), and are returning the NULL pointer if that occurs. That means you need to update all the callers of gb_generate_enum_strings() to also handle a possible null return value. The fix does a good thing, and your description is correct about what you are fixing. But it should supply more complete context for the change. More below. > Fixes: e65579e335da ("greybus: audio: topology: Enable enumerated control support") > Signed-off-by: Jiasheng Jiang <jiasheng(a)iscas.ac.cn> > --- > drivers/staging/greybus/audio_topology.c | 18 ++++++++++++++++++ > 1 file changed, 18 insertions(+) > > diff --git a/drivers/staging/greybus/audio_topology.c b/drivers/staging/greybus/audio_topology.c > index 1fc7727ab7be..e9f47a1f0d28 100644 > --- a/drivers/staging/greybus/audio_topology.c > +++ b/drivers/staging/greybus/audio_topology.c > @@ -146,7 +146,11 @@ static const char **gb_generate_enum_strings(struct gbaudio_module_info *gb, > __u8 *data; > > items = le32_to_cpu(gbenum->items); > + > strings = devm_kcalloc(gb->dev, items, sizeof(char *), GFP_KERNEL); > + if (!strings) > + return NULL; > + > data = gbenum->names; > > for (i = 0; i < items; i++) { > @@ -654,7 +658,10 @@ static int gbaudio_tplg_create_enum_kctl(struct gbaudio_module_info *gb, > > /* since count=1, and reg is dummy */ > gbe->items = le32_to_cpu(gb_enum->items); > + > gbe->texts = gb_generate_enum_strings(gb, gb_enum); > + if (!gbe->texts) > + return -ENOMEM; > > /* debug enum info */ > dev_dbg(gb->dev, "Max:%d, name_length:%d\n", gbe->items, > @@ -861,7 +868,10 @@ static int gbaudio_tplg_create_enum_ctl(struct gbaudio_module_info *gb, > > /* since count=1, and reg is dummy */ > gbe->items = le32_to_cpu(gb_enum->items); > + > gbe->texts = gb_generate_enum_strings(gb, gb_enum); > + if (!gbe->texts) > + return -ENOMEM; > > /* debug enum info */ > dev_dbg(gb->dev, "Max:%d, name_length:%d\n", gbe->items, > @@ -1032,8 +1042,12 @@ static int gbaudio_tplg_create_widget(struct gbaudio_module_info *module, > csize += offsetof(struct gb_audio_ctl_elem_info, value); > csize += offsetof(struct gb_audio_enumerated, names); > csize += le16_to_cpu(gbenum->names_length); > + > control->texts = (const char * const *) > gb_generate_enum_strings(module, gbenum); > + if (!control->texts) > + return -ENOMEM; > + You can't simply return here. If you look a bit above this, where the call to allocate a control structure is done, you see that a NULL return there jumps to the "error" label, so any already allocated and initialized control widgets get cleaned up before returning. > control->items = le32_to_cpu(gbenum->items); > } else { > csize = sizeof(struct gb_audio_control); > @@ -1181,8 +1195,12 @@ static int gbaudio_tplg_process_kcontrols(struct gbaudio_module_info *module, > csize += offsetof(struct gb_audio_ctl_elem_info, value); > csize += offsetof(struct gb_audio_enumerated, names); > csize += le16_to_cpu(gbenum->names_length); > + > control->texts = (const char * const *) > gb_generate_enum_strings(module, gbenum); > + if (!control->texts) > + return -ENOMEM; > + You have basically the same issue here. You can't just return, you must do some cleanup too. -Alex > control->items = le32_to_cpu(gbenum->items); > } else { > csize = sizeof(struct gb_audio_control); >

3 years, 5 months

Re: [greybus-dev] Addition of config USB_HSIC_USB3613 soon?

by Greg Kroah-Hartman

On Wed, Dec 22, 2021 at 10:54:41AM +0100, Lukas Bulwahn wrote: > Dear Vaibhav, dear Johan, dear Alex, dear Greg, > > I have seen that the greybus arche driver has been under heavy > development in 2016 and 2017 with some further clean-up in 2019. > > However, so far, the config GREYBUS_ARCHE for this driver still > depends on the out-of-tree config USB_HSIC_USB3613, with a proper > exception made for compile testing (with COMPILE_TEST). > > Will this USB_HSIC_USB3613 config and driver still be added in the > mainline kernel in the near future, so that the config dependencies > are consistent in mainline? Do you have this hardware? If so, we can add the driver, but given that I did not think the chip ever actually shipped, it didn't make much sense. > Or, are the further out-of-tree additions still maintained for the > current kernel and will stay out of tree? Is this arche driver not > needed anymore and can be dropped? Do you want to drop it as it is causing problems for you? It's a good example driver for those wanting to create a greybus host controller driver so it's nice to have in the tree, unless you have a different one that should be merged instead? thanks, greg k-h

3 years, 5 months

Re: [greybus-dev] [PATCH] Header line: Coding style fix

by Mark Greer

On Fri, Dec 17, 2021 at 11:34:08AM -0300, Jorge Eduardo Fermino Oliveia Silva wrote: [Note: I am traveling for the next week so I won't be very responsive.] Hi Jorge. Before we get to the platch please remember that you should send all Greybus patches to greybus-dev(a)lists.linaro.org and linux-kernel(a)vger.kernel.org. I will add them in now and leave all of the context so other can see what you sent. > Solve CHECK: Lines should not end with a '(' > > Reported-by: Jorge Eduardo Fermino Oliveia Silva <jorgeubermensch(a)gmail.com> > Signed-off-by: Jorge Eduardo Fermino Oliveia Silva <jorgeubermensch(a)gmail.com> > --- > drivers/staging/greybus/audio_manager_private.h | 8 ++++---- > 1 file changed, 4 insertions(+), 4 deletions(-) > > diff --git a/drivers/staging/greybus/audio_manager_private.h b/drivers/staging/greybus/audio_manager_private.h > index 2b3a766c7de7..a17f09a19014 100644 > --- a/drivers/staging/greybus/audio_manager_private.h > +++ b/drivers/staging/greybus/audio_manager_private.h > @@ -12,10 +12,10 @@ > > #include "audio_manager.h" > > -int gb_audio_manager_module_create( > - struct gb_audio_manager_module **module, > - struct kset *manager_kset, > - int id, struct gb_audio_manager_module_descriptor *desc); > +int gb_audio_manager_module_create(struct gb_audio_manager_module **module, > + struct kset *manager_kset, > + int id, > + struct gb_audio_manager_module_descriptor *desc); > > /* module destroyed via kobject_put */ The part you're removing has all of the parameters at the same indentation level and what you adding has them at two different indentation levels so I'm not sure this is a step forward. Since the kernel coding style doesn't address this specific case, AFAICS, I would leave it as is despite the complaint. If others disagree then go ahead as I really don't care much either way. Mark --

3 years, 5 months

← Newer
1
...
22
23
24
25
26
27
28
...
58
Older →

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

greybus-dev