- 22 Aug, 2014 2 commits
-
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
- 07 Aug, 2014 1 commit
-
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
- 28 Jul, 2014 1 commit
-
-
Jose Jimenez authored
This is needed because another project builds this file for the lm32, which is big-endian. This makes no effect on the PC. Signed-off-by: Alessandro Rubini <rubini@gnudd.com>
-
- 16 Jul, 2014 6 commits
-
-
Federico Vaga authored
In order to avoid spurious interrupts from previous running, disable all interrupts before registering the IRQ handler. Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
Federico Vaga authored
This prevent confusion while reading the code because, for example, a question: why initialize both timer and tasklet if we can use only one, and we cannot change the configuration at run-time? fd_irq_init: setup_timer only if the user requires timer mode, otherwise initialize tasklet Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
Federico Vaga authored
Explain why the handler doesn't clear the interrupt in EIC_ISR even if the documentation regarding this register is saying something different. Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
- 14 Jul, 2014 2 commits
-
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch>
-
- 02 Jul, 2014 1 commit
-
-
Federico Vaga authored
-
- 12 Jun, 2014 7 commits
-
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch> Acked-by: Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch> Acked-by: Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
The Kbuild writesin 'src' the correct path to the source code automatically. This avoid us to explicitly write 'M' variable Signed-off-by: Federico Vaga <federico.vaga@cern.ch> Acked-by: Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Signed-off-by: Federico Vaga <federico.vaga@cern.ch> Acked-by: Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
these are not a kernel tools, so we do not need use kbuild enviroment. Use a custom environment to avoid confusion. Moreover, add EXTRACFLAGS to allow people to cross compile between 32bit and 64bit on x86 architecture make EXTRACFLAGS="-m32" make EXTRACFLAGS="-m64" Signed-off-by: Federico Vaga <federico.vaga@cern.ch> Acked-by: Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Without the 'gitmodules' rule you cannot automatically clone the submodules Signed-off-by: Federico Vaga <federico.vaga@cern.ch> Acked-by: Alessandro Rubini <rubini@gnudd.com>
-
Federico Vaga authored
Actually, it reverts c92af50f and pick the files directly from the fmc-bus project Signed-off-by: Federico Vaga <federico.vaga@cern.ch> Acked-by: Alessandro Rubini <rubini@gnudd.com>
-
- 09 May, 2014 1 commit
-
-
Tomasz Wlostowski authored
-
- 14 Apr, 2014 2 commits
-
-
Tomasz Wlostowski authored
-
Tomasz Wlostowski authored
Reason: zio block the channel if the raw_io function returns an error (possibly a bug in zio user trigger handling).
-
- 10 Apr, 2014 6 commits
-
-
Tomasz Wlostowski authored
This fixes a potentially very dangerous behaviour: previously the output was defaulted to channel 1 and running fmc-fdelay-pulse with no parameters by unaware user would result with generation of a 10 Hz waveform on output 1.
-
Tomasz Wlostowski authored
-
Tomasz Wlostowski authored
-
Tomasz Wlostowski authored
By default, outputs are disabled and programmed to 600 ns delay, 250 ns width, 1x repeat and 1 us period. This patch allows safe read-modify-write operations.
-
Tomasz Wlostowski authored
This is a hotfix for the COHAL/FESA guys: the current driver/lib does not decouple enabling/disabling the output and setting its mode (pulse/delay). Previously, disabling the output was leading to loss of its configuration. This patch allows to retain it under the condition that the output is only switched between DISABLED and DELAY modes.
-
Tomasz Wlostowski authored
kernel/fd-zio.c: fix incorrect readout of repeat count if output configured in infinite repetition mode
-
- 03 Apr, 2014 2 commits
-
-
Tomasz Wlostowski authored
-
Tomasz Wlostowski authored
submodules: updated zio to include the __devinit patch (fixes build issues on newer kernels with CONFIG_SPI enabled)
-
- 01 Apr, 2014 3 commits
-
-
Tomasz Wlostowski authored
-
Tomasz Wlostowski authored
-
Tomasz Wlostowski authored
-
- 02 Apr, 2014 1 commit
-
-
Tomasz Wlostowski authored
-
- 25 Mar, 2014 2 commits
-
-
Tomasz Wlostowski authored
-
Alessandro Rubini authored
The should pick other repositories from ../ only if both use and them are submodules of ../ . Only checking that ../zio is a directory is not enough, because it can be a separate clone (old version, for example) or worse an intermediate directory that doesn't host headers. A user reported painful build errors for this reason. Signed-off-by: Alessandro Rubini <rubini@gnudd.com>
-
- 24 Mar, 2014 3 commits
-
-
Alessandro Rubini authored
This directly affects tools/fmc-fdelay-status, although the tool is not changed. Signed-off-by: Alessandro Rubini <rubini@gnudd.com>
-
Alessandro Rubini authored
Signed-off-by: Alessandro Rubini <rubini@gnudd.com>
-
Alessandro Rubini authored
TDC user offset should be applied to all timestamps. The code, however, didn't do that properly, and fast_tdc mode was missing it. Now user_offset is applied where the hardware offset is, like it should have since time zero. This should have no effect, as user_offset is an option nobody used so far, as far as we know. Signed-off-by: Alessandro Rubini <rubini@gnudd.com>
-