Age | Commit message (Collapse) | Author | Files | Lines | |
---|---|---|---|---|---|
2016-08-01 | mbim-codegen: fix message name in generated code doc for parse functions | Dan Williams | 1 | -2/+2 | |
At the point where the codegen happens, {name} is the name of the last function argument, not of the message. Also fix up the documentation to say "Parses and returns parameters" instead of "Create a new request" since parsing doesn't create any requests. New: * Parses and returns parameters of the 'Signal State' notification command in the 'Basic Connect' service. Old: * Create a new request for the 'ErrorRateThreshold' notification command in the 'Basic Connect' service. | |||||
2016-07-05 | Avoid affecting the padding behavior for other services than QMI | Emil Ljungdahl | 2 | -8/+10 | |
2016-07-05 | QMI via MBIM | Collin McMillan | 2 | -8/+8 | |
Signed-off-by: Bjørn Mork <bjorn@mork.no> | |||||
2014-07-11 | build: update copyright years and authors list | Aleksander Morgado | 4 | -5/+5 | |
2013-06-24 | mbim-codegen: handle reading/writing byte-arrays with shared length field | Aleksander Morgado | 2 | -20/+41 | |
2013-06-24 | mbim-codegen: check message type when parsing | Aleksander Morgado | 1 | -0/+23 | |
2013-06-24 | mbim-codegen: fix offset after reading 'ref-byte-array' | Aleksander Morgado | 1 | -1/+1 | |
2013-06-24 | stk: add support for the STK service | Aleksander Morgado | 2 | -9/+21 | |
2013-06-23 | mbim-codegen: add support for fixed-sized byte arrays | Aleksander Morgado | 2 | -9/+62 | |
2013-06-23 | mbim-codegen: rename 'byte-array' format to 'unsized-byte-array' | Aleksander Morgado | 2 | -15/+15 | |
We only used this format type in the Service Activation command, and the format did indeed not have a specific size field for the byte array. | |||||
2013-06-23 | mbim-codegen: allow empty command name | Aleksander Morgado | 1 | -5/+9 | |
For the case when there is a single command in the service (e.g. USSD) | |||||
2013-06-23 | mbim-codegen: fix byte-array handling | Aleksander Morgado | 1 | -1/+1 | |
2013-06-21 | mbim-codegen: support setting arrays of struct references | Aleksander Morgado | 1 | -2/+26 | |
2013-06-21 | mbim-codegen: handle optional fields flagged with 'available-if' in setters | Aleksander Morgado | 1 | -19/+33 | |
2013-06-18 | mbim-codegen: include service name in methods if not the 'Basic Connect' service | Aleksander Morgado | 1 | -1/+4 | |
2013-06-18 | mbim-codegen: handle 'ref-byte-array' type fields | Aleksander Morgado | 2 | -17/+46 | |
Which are byte-arrays specified as offset+length. | |||||
2013-06-18 | mbim-codegen: handle optional fields flagged with 'available-if' | Aleksander Morgado | 1 | -15/+58 | |
A given MBIM message may contain fields which can be processed in different ways depending on the value of another field. This is the case for e.g. "SMS Read" Query operations, where the list of messages may be formatted as either PDU or CDMA, depending on the value of the 'Format' field. So, we now support to have fields configured like this: { "name" : "Something", "format" : "string" , "available-if" : { "field" : "ConditionField", "operation" : "==", "value" : "0" } }, { "name" : "OtherSomething", "format" : "string" , "available-if" : { "field" : "ConditionField", "operation" : "==", "value" : "1" } }, Meaning that: * Field "Something" will be given if "ConditionField" == 0 * Otherwise, Field "OtherSomething" will be given if "ConditionField" == 1 Note that in the message parser, BOTH fields will be provided, but only one will contain valid data. | |||||
2013-06-18 | mbim-codegen: handle 'byte-array' fields in Structs | Aleksander Morgado | 1 | -1/+24 | |
2013-06-18 | mbim-codegen: don't setup reader offset if no fields in message | Aleksander Morgado | 1 | -2/+6 | |
2013-06-06 | mbim-codegen: support new 'byte-array' type | Aleksander Morgado | 1 | -8/+41 | |
2013-06-06 | libmbim-glib: fix reading strings within structs | Aleksander Morgado | 2 | -4/+4 | |
https://bugs.freedesktop.org/show_bug.cgi?id=65452 | |||||
2013-05-21 | mbim-codegen: fix generation of struct _free() methods | Aleksander Morgado | 1 | -1/+2 | |
2013-05-21 | docs: setup sections for commands | Aleksander Morgado | 3 | -13/+39 | |
2013-05-21 | mbim-codegen: generate section for all service specific data types | Aleksander Morgado | 2 | -0/+37 | |
2013-05-21 | mbim-codegen: require a 'Service' entry in the data file | Aleksander Morgado | 1 | -0/+6 | |
2013-05-21 | mbim-codegen: only generate _free() and _array_free() methods when needed | Aleksander Morgado | 3 | -33/+82 | |
2013-05-21 | mbim-codegen: document generated structs | Aleksander Morgado | 1 | -0/+46 | |
2013-05-21 | templates: include documentation for the _get_string() methods in the errors | Aleksander Morgado | 1 | -0/+8 | |
2013-04-21 | mbim-codegen: huge refactor | Aleksander Morgado | 13 | -836/+761 | |
I really don't want to split this change into different commits... The overall changes are: * Skip using Value classes, it just complicates the logic a bit here. MBIM is not as complicated as QMI, so it's fine to define all the logic of the generation into 2 main parts: Message creation/parsing and Struct types. * Support for IPv4 and IPv6 values and arrays. If any of the MBIM protocol designers reads this, be aware that I've been cursing you all in the past few days. If you decided to reference IPs as offsets, why the hell don't you do the same when the values are defined within a struct? * Implement the 'IP configuration' message handling * Support for appending structs and arrays of structs into messages. Not fully implemented yet, anyway. | |||||
2013-04-21 | mbim-codegen: don't treat 'IP' as a CamelCase substring | Aleksander Morgado | 1 | -1/+2 | |
2013-04-21 | mbim-codegen: implement guint64 support | Aleksander Morgado | 4 | -0/+73 | |
2013-04-21 | mbim-codegen: allow UUIDs in 'get' containers | Aleksander Morgado | 1 | -0/+5 | |
2013-04-21 | mbim-codegen: allow UUIDs in 'set' containers | Aleksander Morgado | 1 | -0/+1 | |
2013-04-21 | mbim-codegen: fix generation of the notification parsers | Aleksander Morgado | 1 | -4/+4 | |
2013-04-21 | mbim-codegen: allow reading single structs from messages | Aleksander Morgado | 3 | -21/+54 | |
2013-04-21 | libmbim-glib: rename 'INDICATION' enum to 'INDICATE_STATUS' | Aleksander Morgado | 1 | -1/+1 | |
Keeping the same name as in the docs. | |||||
2013-04-21 | mbim-codegen: properly define the 4 types of messages that we may have | Aleksander Morgado | 2 | -80/+49 | |
MBIM defines 4 types of messages: * Query (host->function) * Set (host->function) * Response (function->host) * Indication (function->host) In particular, the Response message is used as reply for both Query and Set. | |||||
2013-04-21 | mbim-codegen: let the field names be written in CamelCase, as in the docs | Aleksander Morgado | 3 | -15/+17 | |
Write all field names in CamelCase, to match the name given in the docs. | |||||
2013-04-21 | mbim-codegen: skip requiring transaction ID in command request builders | Aleksander Morgado | 1 | -7/+4 | |
2013-04-21 | mbim-codegen: simplify API | Aleksander Morgado | 11 | -210/+276 | |
Instead of per-variable getters, just dump a parser() method which reads all the requested variables from the message. Same logic for the message creators, passing all variables altogether. | |||||
2013-02-02 | mbim-codegen: split Value objects | Aleksander Morgado | 10 | -111/+556 | |
2013-02-02 | mbim-codegen: read 'struct' and 'struct-array' fields | Aleksander Morgado | 1 | -12/+48 | |
2013-02-02 | mbim-codegen: add new 'Struct' type generation | Aleksander Morgado | 4 | -18/+270 | |
2013-02-02 | libmbim-glib: implement 'uuid' type reading | Aleksander Morgado | 1 | -2/+8 | |
2013-02-02 | mbim-codegen,libmbim-glib: improve general array and variable reading | Aleksander Morgado | 2 | -22/+91 | |
In the general variable reading part, we no longer need specific methods for 'command' or 'command-done' message types. Also, when reading arrays, we will specify separately the variable with the array size and the variable with the actual array contents. Main reason for this is that it seems they don't have to be one after the other. The arrays of guint32 will be '0' terminated. This is useful for parameters where '0' is not an expected value. In general, the size of arrays of guint32 values is given in the output 'size' variable when the array is read. | |||||
2013-02-02 | mbim-codegen: setup code generator | Aleksander Morgado | 7 | -1/+626 | |
Setup code generation for the request creator and the response getters. | |||||
2013-01-21 | libmbim-glib: setup generation of enum types | Aleksander Morgado | 3 | -1/+154 | |
2013-01-21 | libmbim-glib: setup core, status and protocol error types | Aleksander Morgado | 5 | -0/+127 | |