- 27 Aug, 2015 4 commits
-
-
Nishanth Menon authored
DRA7XX is a superset of DRA74x DRA75x and DRA72x.. that is not really true for these macros.. Lets just use DRA74x instead to keep things simple. Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Seems fgets needs to have return value checked, but our existing logic seems to have no need for it, so cast to void to make gcc 5.1 happy. Fixes: arch/arm/mach-omap/omap4/uc_audit44xx.c:887:2: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] arch/arm/mach-omap/omap5/lib54xx.c:500:2: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] arch/arm/mach-omap/omap5/lib54xx.c:501:2: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/cpuinfo.c:1576:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/cpuinfo.c:1704:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/cpuinfo.c:1707:4: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/cpuinfo.c:303:2: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:1110:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:1170:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:450:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:639:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:782:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:840:2: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:853:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:872:2: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:884:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:906:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:961:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] common/trace.c:984:3: warning: ignoring return value of 'fgets', declared with attribute warn_unused_result [-Wunused-result] Reported-by:
Robert C Nelson <robertcnelson@gmail.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Stop confusing gcc 5.1. Fixes build warnings: common/mem.h:59:14: warning: inline function 'mem_phys2virt' declared but never defined inline void *mem_phys2virt(void *addr); Reported-by:
Robert C Nelson <robertcnelson@gmail.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Dont keep inlines in public functions, just make them static inlines Confuses the heck out of gcc 5.1 Fixes: common/mem.h:59:14: warning: inline function 'mem_phys2virt' declared but never defined inline void *mem_phys2virt(void *addr); ^ In file included from omapconf.c:49:0: common/lib.h:237:12: warning: inline function 'mhz2khz' declared but never defined inline int mhz2khz(double mhz); ^ common/lib.h:236:15: warning: inline function 'khz2mhz' declared but never defined inline double khz2mhz(int khz); ^ common/lib.h:234:12: warning: inline function 'v2uv' declared but never defined inline int v2uv(double v); ^ common/lib.h:233:15: warning: inline function 'uv2v' declared but never defined inline double uv2v(int uv); Reported-by:
Robert C Nelson <robertcnelson@gmail.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 25 Aug, 2015 10 commits
-
-
Nishanth Menon authored
Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Karthik Ramanan authored
* Fixing the compatibility issue with latest CTT version 1.12.0.4 * Clean up the register dump to make it inline with the CTT tool requirements. The new rd1 has been verified with the latest CTT Tool. Signed-off-by:
Karthik Ramanan <karthik.ramanan@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Statistics collector library does not seem too happy with gcc optimization - especially ones with gcc 4.9. So, disable optimization for the library for now. It is a behemoth of code that was handed over through generations of developers.. kinda a bit finicky to fix. Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
This allows us to debug with native debug tools like gdb/valgrind for issues. While at it, add a default build optimization option - we want to save on space a bit. Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Fix up build warnings popped up due to gcc 4.9 Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Fix up build warnings popped out due to gcc 4.9 Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Fix up build warnings we find with gcc 4.9 Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Fix a bunch of build warnings Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
build warnings thanks to gcc 4.9 -> squash them.. Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Nishanth Menon authored
Fix up build warnings in common directory Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 25 Jun, 2015 1 commit
-
-
Suman Anna authored
The cpuinfo option prints out the PMIC and AudioIC device information alongside the SoC information. This data is gathered by querying the devices using appropriate i2c commands. The I2C bus number being used currently matches the bus numbers from the legacy non-DT boots, and does not match the DT alias numbers from the current DT-based boots. The PMIC and AudioIC info is incorrect as a result, and it also causes additional delays due to i2c timeouts on every omapconf command (especially on OMAP5). Fix this I2C bus number so that the cpuinfo information is printed properly for OMAP4 and OMAP5 SoCs. The behavior on other SoCs is same as before this patch. Signed-off-by:
Suman Anna <s-anna@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 06 May, 2015 2 commits
-
-
Vishal Mahaveer authored
Extend android product name to include KitKat and Lollipop. Signed-off-by:
Vishal Mahaveer <vishalm@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Vishal Mahaveer authored
Index the array based on enum - allows more readable code Signed-off-by:
Vishal Mahaveer <vishalm@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 21 Mar, 2015 6 commits
-
-
Dave Gerlach authored
AM437X can now dump the registers used by the Clock Tree Tool(CTT). Registers can also be written to a file using ./omapconf export ctt [<filename>].rd1. Files written in rd1 format so that the CTT can import the file. Signed-off-by:
Dave Gerlach <d-gerlach@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Dave Gerlach authored
Add support for AM437X in omapconf. CPU should be automatically detected. Basic default operation should the be supported now through omapconf core. Also add cpu_max_speed detection at the same time for core cpuinfo support We compare specific bits against the DEV_ATTRIBUTE register found in the control module for am437x to determine the maximum frequency supported by the SoC. Signed-off-by:
Dave Gerlach <d-gerlach@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Dave Gerlach authored
Previously the CPU type detection logic was a bit messy and also used ONLY the dev_feature register to identify am335x parts. This really isn't correct as am437x also has a dev feature register and there could possibly be a collision between values leading to a part being misidentified once am437x support is added. This patch moves the detection into separate function, one for identifying sitara family devices and one for identifying omap family devices. Also, for sitara devices, the device_id register is used to detect which general SoC is in use and then the dev_feature register is read after to determine is a more specific part (AM3352, AM3354, ...) can be determined, otherwise a general SoC name is used (AM335X). Signed-off-by:
Dave Gerlach <d-gerlach@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Dave Gerlach authored
This patch adds a generic AM_335X SoC to use with device_id detection. We can then use the dev_feature register in a future patch to narrow it down to a more specific part if possible but still identify AM_335X if that fails. Signed-off-by:
Dave Gerlach <d-gerlach@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Dave Gerlach authored
Silence echo to builddate and omapconf_version files to keep pretty print output clean. Signed-off-by:
Dave Gerlach <d-gerlach@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Dave Gerlach authored
This patch makes build output much nicer to look at, similar to kernel build output. Signed-off-by:
Dave Gerlach <d-gerlach@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 03 Feb, 2015 3 commits
-
-
Subramaniam Chanderashekarapuram authored
Ignore cscope.in.out and cscope.po.out in addition to the cscope.out file. Signed-off-by:
Subramaniam Chanderashekarapuram <subramaniam.ca@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Subramaniam Chanderashekarapuram authored
The TPS65917 PMIC is like a scaled down version of the TPS659038 used with DRA72 processor family EVMs. Add support for TPS65917 by defining a the necessary register entries within TPS659038. The voltage regulator mappings are as per DRA72x-evm. Signed-off-by:
Subramaniam Chanderashekarapuram <subramaniam.ca@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Subramaniam Chanderashekarapuram authored
smps_id is not a valid variable in tps659038_vsel_to_uv. This caused a build error when debug flag TPS659038_DEBUG was enabled. Hence fixing the dprintf call. This fixes the compile error seen when building with debug flag enabled Signed-off-by:
Subramaniam Chanderashekarapuram <subramaniam.ca@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 30 Dec, 2014 4 commits
-
-
Nishanth Menon authored
Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Suman Anna authored
AM57xx SoCs have two PRU-ICSS subsystems, each of which supports 64 input events routed to their respective INTCs. Events 32 to 63 for each PRU-ICSS subsystem are also routed through a Crossbar. This patch extends the crossbar dump and audit support to the two PRU-ICSS Crossbar instances. NOTE: The sub-systems are not supported on the DRA7 SoCs, so added a small note to the help text. Signed-off-by:
Suman Anna <s-anna@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Suman Anna authored
AM57xx SoCs have two PRU-ICSS subsystems, each of which has 8 interrupts routed to the Crossbar. Add the interrupt information to the common DRA7 Crossbar data, so that this shows up when doing the crossbar dump. NOTE: The sub-systems are not supported on the DRA7 SoCs, so added a "AM57xx only" to the interrupt information, as the corresponding interrupts are marked reserved in the DRA7 TRMs. Signed-off-by:
Suman Anna <s-anna@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Suman Anna authored
Add the interrupt information to the common DRA7 Crossbar data for couple of crypto interrupts. This is done to decipher the information of the currently listed interrupts in the DRA7 dts file. Signed-off-by:
Suman Anna <s-anna@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 15 Aug, 2014 2 commits
-
-
Edward Shin authored
When -m ma_mpu -p emif1 is selected, the channel retains alldmm (default) as the master. This is wrong since we explicitly requested for mpu as the master for the filter configuration. This occurs because there are only checks for incorrect MA_MPU probe channel configurations, and given none, it assumes a non-MA_MPU config. It needs a check to retain correct MA_MPU configurations. Added the missing checks to fix the issue. Signed-off-by:
Edward Shin <e-shin@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Gilberto Rodriguez authored
Add name2addr functions for CM and PRM and add a lib for AM335X with basic functions to allow 'omapconf read/write' to use the register names as listed in the TRM. Caveat: This only supports PRCM registers for now. Update common lib to call to AM335X find register function. While at it, fix a minor formatting error for prm header. Signed-off-by:
Gilberto Rodriguez <gilrod@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 14 Aug, 2014 1 commit
-
-
Nishanth Menon authored
Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 13 Aug, 2014 6 commits
-
-
Nishanth Menon authored
Currently there is no alignment of data, nor is information provided w.r.t the units of the data being depicted. For example (memtester running in background): omapconf trace bw --m0 ma_mpu --m1 ma_mpu --p1 emif1 --p0 emif1 --m2 ma_mpu --p2 emif1 --m3 dss --p3 emif1 Current Output format: Counter: 0 Master: ma_mpu Transaction: w Probe: emif1 Counter: 1 Master: ma_mpu Transaction: w Probe: emif1 Counter: 2 Master: ma_mpu Transaction: r Probe: emif1 Counter: 3 Master: dss Transaction: r Probe: emif1 delay in us: 1000000 overflow delay in us: 1000000 (iterations=1) accumulation type: 2 iterations (0=infinite): 0 Overflow counter index: DISABLED (overflow delay used) Overflow threshold: DISABLED (overflow delay used) time: 656882508 656849727 32781 -> 211.63 211.63 389.74 0.00 time: 656915293 656882508 32785 -> 151.33 151.33 423.28 0.00 time: 656948082 656915293 32789 -> 222.05 222.05 385.83 0.00 time: 656980873 656948082 32791 -> 226.93 226.93 381.06 0.00 With this change, this becomes: Counter: 0 Master: ma_mpu Transaction: w Probe: emif1 Counter: 1 Master: ma_mpu Transaction: w Probe: emif1 Counter: 2 Master: ma_mpu Transaction: r Probe: emif1 Counter: 3 Master: dss Transaction: r Probe: emif1 delay in us: 1000000 overflow delay in us: 1000000 (iterations=1) accumulation type: 2 iterations (0=infinite): 0 Overflow counter index: DISABLED (overflow delay used) Overflow threshold: DISABLED (overflow delay used) Time Stamp(32KHz ticks) -> Throughput(MB/s) time: End Start Delta -> Counter0 Counter1 Counter2 Counter3 time: 660420813 660388035 32778 -> 753.04 753.04 1387.92 0.00 time: 660453590 660420814 32776 -> 764.95 764.95 1400.03 0.00 time: 660486366 660453590 32776 -> 704.70 704.70 1428.90 0.00 Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Edward Shin authored
The statistics collectors (statcoll) on the L3 NoC allow for hardware monitoring of traffic statistics on the EMIF bus. This patch adds support for accessing the DRA7xx statistics collectors using the trace bw feature. Example: $./omapconf trace bw -m dss Monitors read and write traffic to DSS. $./omapconf trace bw -h Displays help text. Signed-off-by:
Edward Shin <e-shin@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Edward Shin authored
Actually, OMAP5 and DRA7 share the same register address, not OMAP4. Fix the same. Fixes commit a7361b40 ([DRA7][TIMESTAMP] Provide 32k counter register address) Signed-off-by:
Edward Shin <e-shin@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Gilberto Rodriguez authored
Update AM335X modules with the appropriate powerdm and clkdm. Add additional modules to print. Update AM335X clocks for new GFX module. Update common module status show function to adjust for the AM335X. Move a line due to remove a printed line to console for VDD_RTC. Add skip for CEFUSE clkdm. Update AM335X main to call the module status show function. Update logic to print standby status. Signed-off-by:
Gilberto Rodriguez <gilrod@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Gilberto Rodriguez authored
Add AM335X clkdm definitions and functions to be used in omapconf show pwst command. Update common clkdm with new defines. Update to common clkdm to call AM335X clkdm functions. Update clockdm_info struct to accommodate difference between OMAP5 and AM335X (position of main domain status bit). Add new function for the clkstctrl status bit. Signed-off-by:
Gilberto Rodriguez <gilrod@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
Gilberto Rodriguez authored
Add AM335X powerdm definitions and functions to be used in omapconf show pwst command. Update common powerdm header with new defines. Update to common powerdm to call AM335X powerdm functions. Signed-off-by:
Gilberto Rodriguez <gilrod@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-
- 08 Aug, 2014 1 commit
-
-
Gilberto Rodriguez authored
Update the common opp files to support the AM335X opp functions. Update to AM335X main to call the omapconf show opp command. Update help to print the show opp command. Signed-off-by:
Gilberto Rodriguez <gilrod@ti.com> Signed-off-by:
Nishanth Menon <nm@ti.com>
-